Know thy enemy and ye shall be victorious. (ancient Chinese proverb)


Problem Areas

Importance of proper car hygiene

Seal clean sheetmetal well


Shocks / struts design and weld seam wear

Fixing Rust


Honda's CRX is a great car, and it is a mystery why it was cancelled in its prime, to be "replaced" by a car with chassis flex so bad it cracks its windows in vigorous driving and has a tiny fraction of a Rex's vast cargo hauling ability (but also has a fabulous VTEC mill).

Unfortunately, over time there are lots of places where rust can develop. This is true in any car, but in a Honda's case, thoughtful complexities like the two-ply floorpan can make inspection-passing repairs of severe rust a nightmare. Honda has chosen to kill the '98 CRX-type that had been rumored to be slated for production, and the SSM will be a relatively expensive Acura in competition with BMW's Z3 rather than a more MGB GT / MR2-style CRX descendant.  We therefore have a need to preserve our beloved spacious roofed go-karts, if only to point out to Honda and the general world the value of such a good, neglected design.

Consumer Reports shows Hondas (going back to 1989 at least) to be perhaps slightly better for rust than average and a lot better than many other cars. Overall, they are better than Nissan and Toyota. In that first year at least, the worst '89 Honda for rust was the Integra, not the CRX or Civic. Contrary to popular opinion, there really are lots of cars with worse rusting problems than our Hondas.

Problem Areas

Front wheelwells and rain drains

There is a plastic liner for the wheels to sling road crap onto instead of steel, which the rest of the wheelwell is made of. Where the liner meets the fender, road crap can accumulate. Frequently run your fingers along the top of the lip (front and rear wheelwells) to brush out road crap. Wash out well (use as much pressure as possible there) whenever washing the car. Rain water drains down the channel built into the liner; at the bottom, it will tend to spread out everywhere, everywhere being all steel. This will destroy the body at the base of the door hinges, much damage being behind the lower grey plastic cladding (Mk I models at least). The cladding will therefore have to be removed to deal with it all properly; be ready for some ugly sights. This is also where the water comes from that takes out both the seat belt anchor areas on the rocker panels and the floorpan area behind the front wheels.

There are also, on my '85 at least, two mysterious matching rust holes most of the way up the front wheelwell, about 45° up and back from the axle to the rear of the strut sleeve. They aren't in the immediate rain water drainage path, but they are there none the less. I suspect many others have them too, lurking behind undercoating.

Just behind front wheels

This is where Honda's use of two-ply floorpans will kill the car if you don't keep an eye on it. Rust can easily develop between these two layers anywhere in the floorpan. As the floorpan bends up to turn into the firewall, the inside "ply" separates from the lower layer and curves more gradually than the sharper-bending lower layer. If both layers have significant rust perforation, welding sufficient sheetmetal on becomes very difficult; mechanics will tell you it's time to move on to another car rather than try to fix it. Keep after that area, monitoring it maybe once a month depending on weather and use. I had an '81 Civic break one its two floorpan drivetrain mounts loose from inter-layer rust; the two layers simply separated. It would flap freely with torque application, over-stressing the remaining floorpan mount. The 1st gen. CRX puts extra stress on that area with its front suspension torsion bar mounts. That far inboard generally is rust-free, but check on it along with the behind-the-wheels area.

Gaps in windshield / hatch glass moulding

Water will leak in through any gaps in the moulding, which by now (on early CRXes at least) has probably shrunk enough to cause gaps. The water goes down the A-pillars, rusting them from the inside out, and accumulates in the footwells.


Be careful when scraping ice off your windshield. If you accidentally scrape the A-pillars (front roof pillars), the paint will be scraped off, allowing rust to form if the primer is penetrated.  If they are scraped, touch them up quickly and well.

Windshield moulding gaps allow water in around the edge of the windshield, which all flows down through the A-pillars. This makes them, like the rocker panels, aqueducts highly susceptible to internal rust.

Sunroof drainage tubes

There are two drain tubes leading from the sunroof to holes in the floorpan just ahead of the rear wheels. These tubes can easily become clogged, leading to massive sunroof rust since the water can't leave. They also come loose, leading to lots of water in the spare tire well (sometimes outright flooding). This is especially important for folks who've put lots of money into stereo gear sitting in that well. Removal of the drain plugs in the well will prevent well flooding, but not widespread rust along the path of the drain water.  The tubes can be secured with twist-ties and fastened with twist-ties, silicone, epoxy and even expanding foam.

SunroofThe sunroof itself is often diseased with rust, with some folks complaining of rust flakes showering down on them. This will happen if the drain tubes stay plugged.  The sunroof can be filled with rust inhibitor / sealant, like expanding foam or Waxoil (see Rust inhibitors / sealants).

The rails need regular greasing, which both reduces sunroof mechanism wear and seals the rails' metal from the elements.


plastic sheetThere are (supposed to be) plastic sheets in the doors to prevent moisture from leaking into the interior or onto the door speakers. Take the door panels off to inspect the sheets and the door's sheetmetal (it's not that hard to remove). A previous owner could have ripped one or both sheets out for some reason or other (odd speaker installation, lock replacement) or they could be torn (lock replacement following jimmying) and you wouldn't know until the bottom of the door is rusting from the inside out or your door speakers sound odd from water damage.

(Above at left) Plastic sheet is also used to the back of the door, behind the (lifetime Honda warrantee) seat belt retracts.  Notice the easy-to-make tears at the rear of it.

Water can sometimes also get in via the door hinge (bad seal), and possibly even through the hole for the door speaker wires.

Water spread from the bottom of the front rainwater drains will occasionally reach the front bottom of the doors, so if perforation develops there water will flow in, enlarging its new entrance and making its own exits over time.

Ahead of rear suspension mounts

Rocker panels

Just under the lower seatbelt anchor, on the rocker panel, is the second major rust point. Water travels along the inside of the rocker panels from the rain water drain outlet in the front wheelwells, and partly from gaps in windshield moulding. It accumulates just under the seat belt anchor, and will eat away a near-perfect circle of metal surrounding that anchor (I know from experience). Part of the reason it chooses there of all places could be that metal experiences more local stress than other areas, and I learned at Penn State that rust loves stress points.

Rear "footwells"

The rear "footwells" are covered by a layer of thick plastic that will eventually crack like dried hardpan, thereby allowing leaking rainwater from the rocker panels to seep through and live in the cracks until the underlying metal is rusted enough for it to leave. Once these cracks form, pieces should be lifted up for inspection of the metal below (requires either a good flashlight or carpet cutting). I've had a welder tell me the floorpan there is plastic, not metal because of that thick plastic layer. An important factor about that area is that is that it is probably structurally critical, being directly ahead of the trailing arm bushings for the rear beam axle (Mk I model). It's an area to take care of well. Unfortunately, it's the same general area that I found the factory undercoating to be intact without any metal above it; a metal panel had to be welded in.

Top Panhard Rod mount

Autocrossers have apparently found that during hard use of a 1st gen. CRX, rust that forms in the top mount of the Panhard rod (rear suspension diagonal link) can cause the rod to detach, eliminating rear suspension lateral (sideways) stability and surely causing significant damage. Check the mount periodically, at least well before inspection time.

Junction of rear wheelwell and fender

Where the wheelwell curves down to meet the fender is a prime rust area, especially the rear of the seam (see Declaring Total War).

Wheelwell moulding strips

The useless vinyl strips on the wheel well lips trap water, dirt and salt (especially at by the back end of the strip). They should be removed and the inevitable rust there fixed (see Fixing Rust).  Don't bother reinstalling them, since they'll just duplicate the problem later. The sheetmetal at the back end of the strip is especially vulnerable since that's where the most crap accumulates, and there's already a moisture accumulation problem on the other side of the sheetmetal. On some CRX owners' cars rust has extended the entire way around the wheelwell. It all remains invisible without removing panels.


Road crap accumulates behind the metal part above the bumper, especially if one does not have rear mud-flaps. It therefore rusts from behind. The bumper mounting clips on the side, the plastic ones, tend to chip the paint off and the mounting hole rusts out, resulting in a sagging rear bumper. The clips can be replaced with a bolt and should then be sealed with silicone. A carriage bolt (for a square mounting hole) works well because that square fits the plastic rails on the bumper after a little grinding off some of the top of the bolt.

Many CRXes have taken hits, and wherever metal has been bent paint will flake, allowing water in.  This will lead to wholesale rusting of the entire affected bumper & attachment structure for end hits.  Since the front fenders and of course the bumper covers are plastic, all that rust will be invisible.

For anyone who doesn't already know, CRX bumpers are extremely easily removed. There are two long vertical bolts in the bumper, accessible from its underside, that go through the front of the collapsible bumper mounts. Once removed, the bumper just slides off its slots into freedom.

Delaying Rust

Importance of Proper Car Hygiene

As a car is used, road crap (salt, dirt, water, etc.) will work its way into every crevice it can find. If allowed to reside there, it will eventually seep through the paint and feed on the underlying steel. That's why it's very important to well clean a car with lots of water pressure used everywhere. Otherwise, as the sheetmetal develops rust holes, more and more corrosive road crap will find its way in and turn the car into rust-coated steel Swiss cheese from the inside out. Undercoating a car when new will delay crud accumulation in (still clean) joints, seams, etc. but application of it to areas wearing lots of road crap would probably help very little if at all. When I last had rust work done, they strongly recommended that I thoroughly wash the salt and general road crap off the floorpan, so it can't all just sit around enjoying a steel buffet. An '88 Accord owner mentioned how he uses a high-pressure sprayer, and has yet to find rust anywhere on it.

Another factor (but one that can't be easily changed) is the color of the car.  White reflects sunlight while black absorbs it.  The darker the color, the quicker water will evaporate off it.  That is why white cars will generally have a little more outside rust than darker cars.  Keep in mind, though, that most rust still starts on the inside, originating from faulty rain drainage.

I have found that birds prefer white cars for "target" areas more than at least the surrounding asphalt, etc.  My white CRX is like target practice to the local birds.  The ground around it remains unblemished while the white sheetmetal (seldom the glass) is successfully hit.  A caller to the PBS Saturday morning car talk radio show "Click & Clack, the Tappet Brothers" described how he has two identical cars.  One is other white, the other a darker color (red or blue, I don't recall).  The darker car gets spared while the white car is constantly bombarded.  Reasons, unknown.

Seal clean sheetmetal well

If you can prevent oxygen and water from getting to steel free of all road grime, that steel will not rust. If the steel isn't cleaned off first however, whatever sealing applied will trap moisture and corrosive road crap against that steel and it will rust, no matter how much sealing is applied.


I've heard that rubberized undercoating spray will trap moisture against the sheetmetal while standard spray won't, and the shop that does my welding doesn't use rubberized spray. On the other hand, most decent rubberized spray describes itself as "water-displacing." In any case, use undercoating spray often, but take care not to plug drain holes with it!!! Plugged drain holes will cause water to back up until it either finds an opening to pour through, or it will make one. Pour water on the car and watch to see where it exits. That'll tell you immediately if anything is blocked.

Shocks / struts design and weld seam wear

This may seem a little disconnected, but using simple hydraulic shocks and struts instead of gas-charged ones will hasten rust. Without a gas "pillow" in them potholes and other sharp bumps will strain already old weld seams alot more that gas-charged units will because alot more of the impact has to be absorbed by the structure. The peak strain of highway-comfy worn hydraulics to gas shocks / struts over harsh bumps has got to be at least 3:1 (from seat-of-the-pants experience). It doesn't matter how big-Buickish your ride may be on the highway, it's the under-isolated intown potholes that do the damage. The car is after all a monoque / unibody, so it relies on intact weld seams to stay strong and creak-free (and not crumple in an accident). If overstress tears them, corrosive road crap can literally pour in through there. Also, with gas shocks / struts, squeaks and creaks will be reduced and the car be much more settled over bumps (esp. the Mk I's beam rear axle).

Fixing Rust

Preparing the wound area for treatment

For any repair, the wound area should be cleaned thoroughly before treating, and of course first make sure that a rusty hole isn't a rust-enlarged enlarged drain hole (which should first be cleaned and the metal sealed from air and road grime without plugging the hole). With any rust damage, it is important to locate the reason for the damage, or it will just continue on unabated and require a continuous stream of repair work.

Rust flakes should be rubbed off first (a rag works well) and then depending on wound size, use either a Dremel or a normal-size drill with wire brush or sanding disc attachment to finish rust removal. The area must be totally dry and preferably rust-free. You need to remove as much of the rust as you can because it (iron oxide) has no strength or bonding tendencies with the metal it came from (unlike aluminum oxide). Whatever you try to bond to the rust will flake off with the rust itself.  Rust just flakes off, always exposing fresh metal to be eaten until there's nothing left. It is a highly malignant automotive disease.

I found that the nozzle used for inflating basketballs, etc. for an electric portable air compressor worked well for blowing off loose rust particles.  I snipped the end off, behind the side outlet, then crimped it just enough so that I got maximum air pressure out the tip.

Rust treatment then gets applied. Five brands were recommended to me, and probably about a thousand available: Rust Destroyer, Novarox, Rust Check, 3M's Rust Fighter 1 and (British) Waxoyl. There are two  main types: rust convertor and rust inhibitor / sealant.

Rust converters

Rust Destroyer and Rust Check and many, many others, are in the "turns rust into primer" rust converter category. I've used one; there are lots of brands. It stayed clear on good metal and quickly turned to black primer on the ugly areas. Don't expect too much of it, though; it will only neutralize the final, harder to remove, thin layer just above the good metal. The primer that the converting sprays turn it into can be sanded for better final paint adhesion, and has much more strength and bonding ability than rust does.

Naval Jelly is a Vaseline-consistency material that dissolves rust.  After a few minutes of letting it sit, you wash it off.  Washing it off of hard-to-reach interior sheetmetal (more like really small open-air vent holes in most rust cases) is not easy however, unless you don't mind soaking some of the interior (and then having the water sit in its favorite hiding places).  It is also highly caustic, so vinyl gloves should be used.

Rust inhibitors / sealants

Rust Fighter 1 is a yellowish wax in an aerosol can, and two or three light coats will work. It takes awhile to dry, so use only in warm weather and allow lots of drying time between coats. There are probably many other brands out there. A sprayable wax would be perfect for forming a thick barrier for the metal.

WaxoylWaxoil is a combination of wax and oil, making it an excellent rust inhibitor.  A British CRXer poured it through his sunroof, and what had been a localized rust magnet hasn't given him any problems since.  In the UK, waxoil kits are sold in basic auto shops.  The kit consists of 5 liter can of waxoil, a pump and sprayer attachment, and a fill hose.

I recommend using a coat of coarsely sanded (for good bonding), wiped clean, rust convertor "primer" over whatever last small bit of rust you don't have the time/whatever to get rid of, and then using a good, thick, flexible sealant like Rust Fighter 1, Waxoyl or similar to well seal the metal from alien elements.

Alternate Materials & Other Repair

Bondo, Naval Jelly, Rust Fighter 1, WaxOyl and the various Turn Rust Into Black Primer sprays are not the all-inclusive list of successful rust repair materials available.

Also, I've found that's Correless Rust Stabilizer (in their rust prevention section) works well (my currently preferred main anti-rust compound, and no, I don't work from or it's affiliates!).  It has no toxic chemicals in it (it's acid free, unlike naval jelly and others that rely on rust and metal -eating phosphoric acid) and (partly due to that) is claimed to have been used on North Sea oil rigs for the past 30 years.  It's available in both spray and 1qt. paintcan versions.  In really tough areas, spreading fiberglass over top of the wider general area after Rust Stabilizer use to prevent rust appearing nearby is likely a good idea (...fiberglass costs less than Rust Stabilizer).

Expanding Foam

Expanding foam can be used as long as it doesn't block a drain point or let water build up next to it. For instance, if using it at the front wheelwells to stop the rocker panels' aqueduct effect, the foam would have to completely fill the box section from end to end to prevent entry points from being able to form mid-length and build up corrosive road crap behind filler at the ends.  As with any sealant, the metal must be totally dry and clean before application.  It expands from an ultra-sticky, somewhat caustic, expanding paste into a hard, extremely porous mass over 10X of the original in-can volume.

Expanding foam can easily be extremely messy and irritating to work with.  The can must be held, and stored, upside down so the propellant doesn't all escape.  The tube that comes with it pops off from the pressure, leading to ultra-sticky blobs of foam dripping down from both ends of the tube / straw and the can nozzle.  What's really needed is some kind of very small clamp to hold the blasted tube / straw onto the can.  If you can be successful in at least cleaning up after it, it does do an amazing job of sealing off very large volumes in rocker panels, fenders, anyplace really, preventing from water settling in them.  Just don't let water settle on top of it.


Fiberglass is non-porous, fairly easy to work with (even easier than Bondo) and sets hard in a relatively short span of time.  It's being non-porous makes it better than Bondo for anti-rust work.

High-Strength Patch Material

JB Weld patch material has been found to be strong where patch strength is needed.  There's also product called QuikSteel that says it "bonds like epoxy ... hardens like steel," which I haven't tested yet.  Since thin sheetmetal on older Hondas is so hard to weld well, any patch material that will come close to matching the original steel's strength would help minimize new creaks and allow slightly better resale value vs. lots of Bondo.

New Exhaust Holes

For exhaust system leaks I've found cement-like stuff called Muffler Weld (made by Qualco, comes in a flat can) to work well but will eventually start falling back into the muffler as its rust spreads.

Preventing Battery Terminal Corrosion

Many folks, including myself, have found our cars to start ok, but then have the radio and other accessories shut off whenever a significant electrical load is applied, like brake lights, headlights, blower, etc.  The startup current for each is so much more than their running current that a good, low-resistance battery connection is needed to prevent usually temporary alternator overload at low rpm, and electronics damage from voltage spikes at higher rpm.  Applying oil, grease or Vaseline to battery contacts will work for preventing corrosion build-up there since no oxygen can then reach the contacts.

Miscellany - Cleaning Interior Pieces
(you want it to look really good, right?)

Apparently a mix of a little bleach, a little dish washing detergent, and a bunch of hot water applied with a half soft sponge and half synthetic Brillo kitchen sponge works best.  For greasy and oily marks, paint thinner works.  Use mineral spirits, not a lacquer thinner.  It will de-hydrate and de-grease the spot however. One has to use a vinyl protectant to re-hydrate and re-oil it again.

Declaring Total War

Most rust is hidden behind plastic panels, whether they be the outside grey cladding on Mk I models, rocker panel covers, or rear shock tower covers / speaker enclosures. You must remove all plastic panels to properly see and access the rust. The rust below the door hinges is a perfect example; that's hidden largely by the door itself.

Rocker panelRust forms wherever water spends the most time; low, around possible water entry and drainage points. Sheetmetal up high will be nearly perfect while metal surrounding low-level bolt holes will be a catastrophe.

This is the basic initial weaponry you'll need.Basic tools A 10mm socket (at left) can remove most bolts used on a Honda's body.  A file and a good-sized standard screwdriver work very well for scraping out rust, with the latter doing later Bondo mixing duty.  A pair of medium-diameter Philips screwdrivers are needed for plastic cladding screw removal (one long, one stubby).  Spray can lids come in very handy for holding screws, bolts and flanges.  Alot of scraping power is needed; nothing less than a wire brush attachment for a good electric drill will do (it'll need cool-downs).  As mentioned, that also allows the rubbing off of protective metal from the wire brush to the car's sheetmetal.  I used naval jelly, followed much later by rust convertor for maximum terminal overkill.  That was followed the next day by Bondo and expanding foam use (neither in this Day 1 pic).

That was in '97... then I tried Correless Rust Stabilizer (no metal-eating phosphoric acid, unlike naval jelly) and experimented with fiberglass use (both being non-porous).  Except on exterior surfaces, I would these days replace Bondo use with fiberglass use.

This is a little modification I did to make sure all loosenozzledust and debris got cleared out of crevices and impossible to reach crannies.  It's an inflation probe, with the tip cut off behind the sideways hole, and crimped down for maximum exit airspeed (too low and condensation will simply "glue" dust to metal while not removing debris).  I used a decent house-voltage tire inflator; anything cheap wouldn't work hard enough, but would still overheat quickly from the high continuous pressure and low airflow.  Even my Black & Decker got really hot, only partly from sitting out in the sun.

The CRX, Mk I at least,cladding plastic "screw"uses plastic screws and flanges for them to hold the cladding onto holes in the rocker panel (as mentioned above).  If you use even a little too much twisting, or don't push the screwdriver at the screw hard enough, the screw slots (Philips) will be destroyed and you'll have to use pliers to yank it out with the flange (usually just tearing the head off both).  Sometimes the screw won't want to come out anyway, and the only removal method, given their low strength, is to destroy it with its plastic flange.

The bottom flange holes become used by rocker panel water (aqueduct effect) as exits.  That means there will be an ugly circle of rust around each one.

Rocker panel frontThis is what many CRXers will see, the way I did on my '85 CRX DX, upon lower grey cladding removal.  This is the front right jack "hard point" ... more so in the past.  Without treatment it would have crumpled upon future jack use, which has already happened to many younger Hondas.


If the car was wet recently, give it extra time to dry. The rust spots will be wet after the rest of the car is dry simply because that's where the water drains to.

Follow Preparation instructions listed earlier for removing all rust.

After wire brush wheel useWhen grinding, once the paint, rust flakes and then rust dust stop flying from the wire brush wheel (use eye protection!), the wheel will rub a zinc(?) layer onto the surviving actual metal.

Before......and after
What the left rear wheelwell lip looked like before and after going at it with weaponry and then coating what survived with rust convertor

Since water does follow gravity so well, you'll need a brush to efficiently apply gel-type rust convertors, and then inhibitors, or preferably both replaced by Rust Stabilzer, since the rust will largely be in hard-to-spray-evenly wee crevices and sheetmetal folds and junctions.  Sprays work ok, but will drain and drip away from where you want it to stay.

Behind back left speakerInside of back right wheelwell, to the rear behind the speaker, after file, screwdriver and rust convertor use


Decide on what gets the Bondo or fiberglass treatment, and what gets new sheetmetal welded in (followed by good sealing of the new sheetmetal).  Keep in mind that the more that welded sheetmetal is used over Bondo or fiberglass, the stronger the car will be.  Also keep in mind that over the next year or two there's a fair chance that the welded-in panel and even its welds will still be completely intact, but one or more sections of floorpan sheetmetal where the welds had attached to will have receded away from those weld seams due to rust, rendering the welded-in patches a bit of a waste as the automotive cancer of rust turns malignant.

In the case of the rocker panels and behind the rear wheelwells, Bondo proved of no use since the cause of the rust was water being allowed to pool there, both being low areas of the car.  Large volumes needed to be displaced by something light and inexpensive that absolutely will not let water settle and/or seep in.  Expanding foam seemed the only way to go.

Fill 'er up!  :->Behind the back left wheelwell, a few minutes after applying the expanding foam


Sealing the edges of the expanding foam with fiberglass would be a good idea, however possible, since after a couple years this DX _again_ started having signs of interior rust in the wheelwell edges.  I suspect there was leakage between the expanding foam and the wheelwell, and that the expanding foam may have been slightly porous.

The expanding foam could simply be replaced by fiberglass, as long as it's smoothly enough applied to let water run down and away from the wheelwell and not settle in various rough spots.  One major benefit of it would be its being translucent, so any new perforations in the sheetmetal would show through very well when viewed from behind, very unlike expanding foam.

Bondo Use

If there is a dent involved, then if you can hammer out the dent from the inside of the dented metal, do that. Otherwise you have to drill a hole (or holes) into the dent and pull it out. This, like all bodywork, is an art. Pop the dent out to about 1/8" below where it should be (leaving room for body filler).  A place called the "Dent Doctor" uses dry ice and a suction cup to pull the dent back.  If the dent is pulled within only a few days of its creation, the metal may be able to pop back to its original shape with minimal trace of its having been there.

Prepare the damaged area per Preparation. Make sure all rust is gone from metal, or it will spread again. The Bondo can directions call for adding an extra inch of sanding away from the wound edge, which would be advisable with any compound use.

Apply some sort of rust inhibitor (Novarox, Rust Fighter 1, etc). The Bondo can directions don't specify this, and I'm not sure how needed it is or how well Bondo will cling to it, but it worked for a CRX driver on his parent's car long enough ago for any problems to show up. Apparently none did. Rust inhibitor is designed after all specifically to block rust while Bondo isn't.   or better...  Sparingly use some of Eastwood's Rust Stabilizer.  The purpose here is to seal the sheetmetal better than primer alone would, in any case.

If the rust was bad enough to alter the surface contour then apply Bondo to the area to fill any indentations/holes. Cover severe holes with wire screen underneath for something for the Bondo to attach to. Don't take more than a minute to mix and apply it, since it hardens quickly and without much of a transition from liquid paste to more like dried dough.  Wait for it to set up and cure. It's ready for sanding in 15 to 20 minutes.  Fiberglass of course doesn't work for surface work because it doesn't respond well to sanding.

Sand the Bondo surface smooth. You can use a sanding disc on the drill if need be. Use coarse, then fine automotive sandpaper found in auto parts superstores. Start with course grain, and proceed to fine grain then using you hands (get / make a sanding paper block).

"Feather" the edges of the Bondo so it blends with the surrounding surface.

Mask any area you don't want painted, but DO NOT spray up to the masking unless it lies along some trim or something.  Outdoors, overspray ends up everywhere, so apply masking extra-far from the intended paint edge.  Apply primer to area (up to three coats will work).  If need be, apply putty for remaining imperfections, re-sand and re-primer.  If the area is large, you may want to repaint the entire panel to avoid color blending problems.

Wetsand the primer to make it as smooth as you want it.

Apply "matching" paint.  Tremelclad has been recommended repeatedly.  It is better to use urethane-based paint, since enamel is too soft and will scratch easily.  Use several coats.  Try to find the exact match to your make/model, or it will really show. This won't be easy, since there is no way Greek White (to take my Rex as an example) found new in a can will match what on my car has seen sunlight and general whatnot for 12 years. You may have to bring in a paint chip right off your car and match it to whatever you can find. The can directions call for spraying primer and paint right up to the masking, which would guarantee everyone noticing the patchwork by each well-defined edge. Don't spray all the way to the masked edge, unless the edge is at a piece of trim / window / another panel / etc.

If the rest of your car uses it, apply several coats of clearcoat.  To get the best results, wetsand it with 600-grit paper.  This step is really needed in any paint blending area, if using smooth gradations from new to old paint since what's in between amounts to course overspray.  Multiple coats of paint can itself produce a smooth surface, but you won't get that smooth, tightly reflective surface in the color blending area around the wound without smoothening the deliberate overspray via fine-grit wetsanding.

Then apply rubbing compound followed by polishing compound.

That's it.  If done carefully, it looks good (possibly very good) and stops the rust.  Only if you're generous with the paint coats and then wet-sand it all will the new paint be as shiny as the original paint.  If you get lazy in your wet-sanding the paint, it will show by not having as smooth a finish as the factory paint (unless the factory paint is has turned to fine sandpaper from neglect) where you turned down the elbow grease.

Floorpan holes

If you have a hole in your floorpan, if possible, don't just Bondo/fiberglass it or rivet some sheetmetal on for replacement. It won't take any stress at all, thereby causing the remaining floorpan area to have to compensate (and the greater the stress, the greater the rust).  Sheetmetal really should be welded on. Whenever I've had to have floorpan welding done, the shop that does it (referred to by my local independent shop trusted for 30+ years) has complained about how "soda can-like" the metal is in its thinness. They say it's very easy to melt the metal you're trying to weld to, and that few places will touch thin sheetmetal cars for that reason.  And again, the surrounding area, above, below and even the area between the two original layers of the floorpan must be rust-free and kept well-sealed to keep long-term exposure to water and oxygen from making it break away from the patch's weld seams and receding the original floorpan from those weld seams.

If there's enough good metal left to constitute a floorpan in the rusted area, follow Preparation directions. First deal with the source(s) of the water otherwise you'll need to either remove a drain plug, which could let tire splash enter, or drill (and rust seal) a new smaller drain hole. If the damage is too great, new sheetmetal will have to be welded into place, first requiring the removal of rusted edge-of-hole metal. Once the new metal is welded on, then apply Rust Stabilizer or similar, followed by undercoating.


Japanese cars have a reputation for being "just drive it" cars in need of no extra maintenance. That fully applies to Hondas, but older Hondas, like CRXes (esp. Mk I models) use thin sheetmetal which past a decade will develop serious  to fatal rust problems if not attended to. Regular undercoating applications will help a little but won't touch the sources of the rust. Most of a CRX's rust is from leakage from the front rainwater drain outlets, leading to an aqueduct effect in the rocker panels, and from water and crap collecting in the rear wheelwell lip vinyl strips and the wheelwell / fender seam on the other side of the metal. On Si models, loose sunroof drain tubes have on some cars completely flooded the spare tire well. Since the CRX is a minimalist vehicle design, loss of sheetmetal and weld seams will significantly reduce the car's strength. Unchecked, rust grows exponentially until you find you have to get another car really quickly because your CRX isn't present enough to pass inspection.

Compiled by:

Roger Twitchell III
COG Member #21

who had an '85 CRX DX for 4.5 years (147,000 mi. at time of sale), and who now has an '85 CRX Si he drove up here to York, PA from Virginia (basically rust-free!  Yeeaaaahhhh!!!  :->).

from over two yrs. of CRX and Honda List participation, odds and ends, my own car's problems & recent rust work on it, CRX and Honda List subscriber contributions, and lifted from the newsgroup.

My thanks, and I'm sure from all others benefitting from this, to those helpful few who (nearly all unknowingly) helped with this.  I should list all the names (and maybe email addresses), but for one, the contributions go back too far, and many of the contributions were things that could be picked up from many other sources, like books or a buddy at a bodyshop or your mechanic.  Therefore, I extend my thanks mostly to CRX List contributors in general which is where contributions have mostly been from.

For the curious: the camera equipment used (sunroof and WaxOyl pics British imports) was a (pre-1981) Petri FTE with either its 55mm f1.7 normal lens often with one or more magnifying lenses on it, or a Sigma 39 - 80mm f3.5 zoom w/macro capability.  I used 400ASA Kodak, Fuji or Clark Photo Labs film with, initially, a very cheap Mustek 200 optical dpi hand scanner (home-improved), Paint Shop Pro, iPhoto Plus, and AOLpress. More recent picture improvements have reflected use of a comparitively fabulous, newer 30-bit, 300 x 600dpi flatbed (SCSI) scanner.

Comments?  Email me.Mailbox

Return to Index

MidAtlantic CRX Meet '99 / CRX Friday At The Track III

Aircraft Engines:  Climbing the Sky
Guide to Internal Combustion Engines, Automotive Engines
Link so you too can learn How to Drive Like a Moron
Improved Driving Techniques
My Cars: Road-Dancing over the Years
What Urban Sprawl, via modern propulsion, is replacing
Classical VR: Reconstructing Lost Structures
How to photograph