The Online Magazine For Amateur Boat Builders

Oil vs. Latex:

Almost-Scientific Testing
by Rob Rohde-Szudy

Everybody seems to have an opinion, but nobody really has any data. Such is human nature, but painting a boat involves enough effort that I thought it would be nice to have at least a LITTLE data.

The paint industry has a wealth of data available, but precious little of it applies to our use. The boat paint manufacturers test their coatings for use on ocean yachts that live in the water year-round. House paint manufacturers test for the stresses involved in houses. We who build small boats are a very small market with an entirely different set of circumstances that nobody really tests for. We generally dry-sail our boats, and they live in the back yard – like house paint. However, we also drag them over concrete at the ramp and gravel and sand at the beach – not at all like house paint OR live-in-the-water marine paint!

These experiments are designed to approximate the stresses that we homebuilders put our paints through, and provide us at least the beginning of some useful knowledge about this area. More that anything, it is an invitation to other backyard boat builders to become backyard scientists for our mutual benefit. These data only apply to the exact paints tested under the exact conditions tested, but we can all contribute to this pool of knowledge. Just make sure you record all of the conditions so later researchers can interpret them as they will.

I decided that I would set up the experiment to determine which, if any, combination has better resistance to the events that trigger recoating. These are the paint being scratched up, the paint peeling, and the paint fading. The bottom of the boat takes the greatest beating, so I tried to replicate its conditions. I mostly tested for scratching and moisture-related peeling. As a great many people are coating the bottom of their boats with fiberglass in epoxy, I used similarly coated samples.

Here’s the basic outline of what I tested:


  • Latex vs. oil
  • Out-of-the-can vs. thinned
  • Gloss vs. semi-gloss
  • Primer vs. paint on epoxy


  • Plywood w/ glass & epoxy


  • Behr Premium Plus Gloss Exterior Latex, Chocolate Cherry, New
  • Mautz Satin Int./Ext Latex, 23-30White, 1/3 gal., a few years old
  • Glidden Evermore Gloss Ext. Oil, Off White, 1/3 gal., 2 years old
  • ACE Formosa Gray Alkyd Porch Enamel, 1/2 gal., unknown age but probably very old
  • Behr Premium Plus Int./Ext. Latex Hi-Gloss White #8050, New


  • Behr Premium Plus Exterior Latex Primer #436, New
  • Behr Premium Plus Int./Ext. Oil Primer #434, New
  • No Primer


  • Hard abrasion resistance – dragging weighted sample across wet concrete
  • Wet/dry cycle resistance – mud soak and air dry for each step of the other tests


Of the possible permutations, I left a few out so I could have two samples of each combination for a control sample. See below for combinations tested. I chose to leave out combinations of oil over latex, since I think that anyone who is willing to mess with oil for the topcoat would just as soon use it for the primer. I also left out any thinning when coating paint directly to epoxy, as people who want to get it painted and in the water do not put on multiple thin coats. I also left out the satin latex paints over oil primer, retaining only the gloss latex as a comparison with gloss oil.

Samples were lightly sanded with #60 grit and wiped off with damp paper towel, then dried for 30 minutes before coating with primer. No preparation of prime coat to take paint – just a week of drying time. Both coats were applied at room temperature and humidity (64 deg. F, 29% humidity). All samples were cured 4 weeks after last coat. The thinned examples were thinned not quite enough to avoid brush marks – possibly less than I should have thinned them. I was using cheap disposable brushes, which tend to leave brush marks. Primer was flowed on with the grain with no attempt to ‘stretch’ it. (That is, the brush was kept good and wet.) Then it was immediately cross-brushed at a 90 degree angle to mitigate brush marks. (With a good brush this shouldn’t be necessary, but many of us use the cheapies, so that’s what I tested.) Topcoat was put on in the opposite directions. (Again, with a good brush this shouldn’t be necessary.)

Observations During Coating

Courtesy of Duckworks publisher Chuck Leinweber, I had big box of 6x6” squares of ¼” plywood. These were beautifully coated with 6 oz. fiberglass set in epoxy and sanded ready for painting. I could see why Chuck’s boats look so nice.

Primer applied 3/2/03, 64 deg. F, 29%. Paint applied 3/8/03 in similar conditions. Here’s a photo of a representative sample.

Latex paint and primer flowed on more smoothly than oil, and required less dipping of the brush. Needless to say, thinned versions required constant dipping.

The gray porch paint had a THICK layer of skin that required a utility knife to get through before stirring. The off-white oil paint had a skin, but not so horribly thick. The older latex paint and the new paints had no skin. The porch paint was so thick it could not tolerate cross-brushing without marking up like pudding. I thinned it rather too thin and it would barely cover. An interesting test in its own right.


The testing was set up to parallel the habits of cartoppers and trailer sailors – occasional hard use and lots of sitting out in the weather completely neglected. The control samples went through the same abuse, but without being abraded.

Thus, after coating and curing, I dragged 3 times them across wet concrete for about 20 feet with about 30 lbs on each sample. Results were about the same for all samples – light surface scratching all over with occasional gouges that took off all paint.

Then I soaked them in muddy water for a couple days and spread them out face down in the sun.

About a month later I flipped them face-up. All still looked about equally scratched and muddy.

Then I put them all back in their cardboard box and let the weather work on them for a few months. They appeared the same after this treatment, so I put them in a plastic garbage bag (wet) for a year. Actually I did this because I had to get them out of the way in a hurry, but then I completely forgot about them for about a year. (No comments!) The bag was stored outside and leaked, so they stayed wet. Again, no real differences. The plywood did not look very happy about this treatment, but again, no differences emerged in the paint.

Here’s a representative sample:

I cleaned and dried the sample so we could actually see what’s going on with the paint.


Within this test, oil and latex came out equal. This is limited to abuse by abrasion and humidity, as the solar radiation I subjected the samples to was relatively minor. Sadly, the samples were later lost in the process of moving, precluding further testing.

It does appear that for our purposes scraping resistance is the primary difficulty with housepaint of any kind. This hypothesis is especially supported by the fact that the samples without primer did just as well. Apparently a scraping force powerful enough to lift paint is also powerful enough to lift paint and primer. The main benefit of primer seems to be that it was easier to get coverage without sagging in two coats, and that primer is cheaper than gloss paint.

Questions Raised

1. Would differences emerge over longer-term testing? I think 3 years is the most we would care about, as after 3 years most of us have beaten up the paint enough that it will need recoating anyhow. (When cartopping I think it gets one season beats it up at least that much if you use your boat much at all.)

2. Would some type of marine paint improve abrasion resistance enough to justify its cost? To me this would mean adding 1 or more seasons before recoating.

3. How would the same paints fare coated on plywood with no fiberglass for topsides applications? Would unprimed samples fare worse without the epoxy filling the wood’s grain? Would checking then become the limiting factor? For that matter, how do you keep plywood from checking? (Had I been really smart I would have coated both sides of the original samples, getting two tests from each. I wasn’t.)

4. This test in no way applies to boats that live in the water. My feeling is that if they can afford the expense inherent in that type of boat, they can afford real marine paint. Or they can write an article on how housepaint measures up in that environment. I am reminded of Bolger’s remark that you should build a boat big enough to live on or small enough to take with you.

5. This test also had nothing to do with saltwater. Sorry, coastal folks. Most of us live on the muddy “fresh” stuff.

Combinations Tested

Not thinned
Not thinned