Water-Based vs. Oil-Based Polyurethane: 44% Solids vs. 28% in the Data Sheets
Water-Based vs. Oil-Based Polyurethane: 44% Solids vs. 28% in the Data Sheets - A step into the workshop.
Stop Wasting Wood
Generate clear cut diagrams instantly with our new Cutlist Optimizer. Plan your cuts to make your next build easier.
Every finishing argument I've had at a lumber counter runs on vibes. Oil is warmer, water is clearer, oil is tougher, water is safer, and nobody in the conversation has read a technical data sheet in their life. So I pulled the sheets and read them side by side. The first thing that jumps out has nothing to do with which finish is better. It's what the manufacturers are willing to put in print.
The Sheets Don't Even Play the Same Game
Rust-Oleum's data sheet for Varathane Water-Based Floor Polyurethane publishes a full lab battery: pencil hardness of 3H under ASTM D3363, 50 mg loss on Taber abrasion (ASTM D4060, CS-17 wheel, 1000 g load, 1000 cycles), 4.0 kilograms to rupture on the Hoffman scratch test, 120 in-lbs direct impact and 90 reverse. It even prints KCMA chemical ratings, where vinegar, ketchup, coffee and olive oil all score a clean 10, and 100-proof alcohol and mustard both drop to a 7.
Rust-Oleum's oil-based floor sheet publishes none of it. No hardness, no abrasion, no impact, no chemical resistance, just physical properties and how to brush it. Minwax's oil-based sheet is the same story. Its only ASTM citations are sheen, dry time, and VOC content.
That matters, because Bob Flexner wrote in Woodshop News in March 2018 that oil-based poly is more scratch-resistant than water-based. Pros on WoodWeb said the same in July 2000, but that thread is 26 years old and belongs to the history of the category, not to the cans on the shelf now. Flexner gives no numbers, and neither does any oil-based sheet in the stack. So the trade consensus points one way and the only published lab data belongs entirely to the water-based products. Neither of those is a controlled test. If somebody tells you oil wins on Taber, they're comparing a number to a blank space.
44 Against 28
Here's the cleanest comparison available, and it's the one that actually explains the two finishes. Same manufacturer, same product category, same 275 g/L VOC ceiling.
Rust-Oleum's sheet for the oil-based Varathane Polyurethane Floor Finish gives 43.7 to 44.0% solids by volume and a flat 275 g/L VOC. The separate sheet for Varathane Water-Based Floor Polyurethane gives 27.2 to 27.9% solids and puts its VOC at less than 275 g/L. Roughly 1.6 times the solids in the oil can, under the same regulatory ceiling. Everything else follows from that one number.
That's why Rust-Oleum recommends two coats of the oil and a minimum of three of the water-based. Run the arithmetic yourself on their own figures, and note this is my math on their sheets, not a number they publish: oil at 2.0 to 3.5 mils wet times about 44% solids lands at 0.88 to 1.54 mils dry, which matches its stated 1.0 to 1.5 mil dry film. Water at 3.5 mils wet times 27.9% comes out around 0.98 mils. Two oil coats put down roughly 2.0 to 3.0 mils total. Three water coats put down about 2.9. Comparable film, if you apply the third coat.
Worth flagging: the water-based sheet claims 1.0 to 2.0 mils dry per coat, but its own wet-film and solids figures can't produce the top of that range in one pass. The same document also gives recoat as 2 hours in one place and 4 hours in another. I'm not resolving that in the finish's favor. I'm telling you the sheet argues with itself.
The calendar is where you feel it. The water-based Varathane sheet puts recoat at 4 hours, so three coats fit in one Saturday. The oil-based finish has its own sheet, and that one says 12 hours to recoat and 7 days to full cure, so two coats eat a weekend. Minwax's furniture-grade products run the same pattern: Fast-Drying Polyurethane recoats at 4 to 6 hours, its Water Based Polyurethane at 2. But that Minwax waterborne only builds 0.6 mil of dry film per coat, so three coats gets you 1.8 mils, thinner than the 2.0 to 3.0 mils that two coats of Varathane's oil lay down. Minwax says outright that three coats are required for maximum durability. General Finishes wants 7 to 10 days before light use on its High Performance topcoat and 21 days to full cure. Fast to recoat is not the same as fast to done.
And here is the number that should end the argument. USDA Forest Products Laboratory, Wood Handbook Table 15–3, moisture-excluding effectiveness of oil-modified urethane varnish on ponderosa pine:
| Coats | 1 day | 7 days | 14 days |
|---|---|---|---|
| 1 | 55% | 10% | 2% |
| 3 | 90% | 64% | 44% |
| 6 | 93% | 76% | 62% |
One coat is down to 2% at two weeks, and FPL cautions that its values below 20% are rough approximations only. Three is a real barrier. Six is better still. That table has no waterborne entry, so don't read a water-based number into it, but the shape of the curve is the lesson: coat count does far more work than carrier chemistry. FPL is blunt that "no coating is entirely moisture proof," and that an effective coating "merely slows absorption or desorption of moisture; it cannot change the equilibrium moisture content." Film-formers work "provided the films are thick enough." That's the whole game.
Water-Based Doesn't Mean Clear
The National Wood Flooring Association's public line is that oil-based finishes are amber and water-based are normally clear. Fine as far as it goes. Then read Minwax's own water-based sheet, which lists its resin as an "Oil Modified Polyurethane Dispersion" and warns that "slight ambering may be experienced when applied over light-colored surfaces." Same warning as their oil-based sheet. Minwax points you at Polycrylic for light woods instead.
Water-based describes the carrier. It does not describe the resin. Varathane's water-based floor product is an acrylic urethane and its sheet says it dries crystal clear, though it still warns about discoloration over white or pastel paints. Two water-based cans, two different answers. Check the resin line on the data sheet, not the marketing on the front label.
Separately, your cherry or walnut is going to change color no matter what you put on it. FPL states that indoor wood color change comes from visible light rather than UV, and that most of it happens within 2 to 3 months.
Your Shop Is the Adversary
Every sheet specifies a window, and they're tighter than a garage in Austin can hold. Both Varathane floor products want 55 to 90°F with humidity under 85%. Bona Traffic HD wants 65 to 80°F and 40 to 60% relative humidity. By July my shop clears the top of that temperature range before noon, and 40 to 60% humidity is not something I can hold in there. Greg Williams, writing in Woodshop News in 2013, put it plainly: humidity has the greatest effect on waterborne evaporation, and don't go below 65°F. Misco's technical bulletin on minimum film-formation temperature is about floor polish rather than wood finish, so take it as mechanism only, but the principle is real. Below a certain substrate temperature a waterborne film doesn't coalesce, and a concrete slab runs colder than the air above it.
Then there's grain raise, the water-based tax nobody budgets. Rust-Oleum builds it into the instructions: if the grain raises, sand smooth before the final coat. The oil sheet has no such line. WoodWeb's 2014 discussion is useful here. Dull paper makes it worse because the grit pushes fibers over instead of cutting them, stock above 7.0% moisture content is more susceptible, and contributors sand 400 to 600 between coats. Minwax specifies 220-grit between coats after a 2-hour dry.
Do not reach for steel wool. FPL warns that steel particles left in the wood cause iron stains later, and Rust-Oleum's water-based sheet says the same thing about rust developing under the finish. Two independent sources, one answer.
One more prep note: FPL classifies oak, ash, walnut, hickory, elm and mahogany as large-pore hardwoods that need filler for a smooth gloss varnish film, and maple, cherry, beech and poplar as small-pore. That decision happens before you open either can.
Two Ways to Get Hurt
Neither of these is the safe option. They're different hazards.
Oil is a fire risk, and not a theoretical one. Minwax's oil-based sheet carries a boxed DANGER that rags, steel wool and sanding residue "may spontaneously catch fire if improperly discarded," and instructs you to put them immediately in a sealed, water-filled metal container. It adds that over-application, low temperature and high humidity increase cure time and the combustion risk, and that sanding dust between coats should come off dry and powdery. Gummy means stop and wait. The National Park Service documents one of its own crews setting off a linseed-oil rag fire and cites an NFPA estimate of 14,070 spontaneous-combustion fires a year. Varathane's oil floor finish has a flash point of 105°F. The water-based version is above 212°F. In a Texas garage in August, that difference is not academic.
Water-based has its own problem. Minwax's water-based sheet, revision 11/19/08, states it contains dimethylethanolamine and 1-methyl-2-pyrrolidone, with a warning about reproductive effects and harm to a developing fetus. That's a can I'm not opening with family in the shop. Meanwhile Bona lists only water, polymeric resins and amorphous silica and holds GREENGUARD Gold certification to UL 2818–2022. Low-VOC and non-toxic are not the same claim, and formulations inside the water-based category vary wildly.
The Money Ladder
Prices read off retailer pages on 26 July 2026. Ace Hardware listed Minwax oil-based satin at $54.99 a gallon and $17.99 a quart, and Minwax Ultra water-based satin at $74.99 a gallon. Woodcraft listed a pint of General Finishes High Performance at $29.99; Oak Tree Supply listed the gallon at $108, sold out. Floor Mechanics listed Bona Traffic HD with hardener at $152.95.
Roughly 40% more per gallon to go water-based at the consumer tier, before you account for the extra coat. Pairing each price only with that product's own coverage figure, my arithmetic: Minwax oil at 600 to 700 sq ft/gal works out to about $0.08 to $0.09 per square foot per coat, so call three coats a quarter. General Finishes at 500 to 600 sq ft/gal runs $0.18 to $0.22 a coat, about $0.59 for three. Bona at 350 to 400 sq ft/gal is $0.38 to $0.44 a coat, roughly $0.82 for two.
Skip the two-component. Bona Traffic HD is labeled for professional use only, and the sheet explains why in its own terms: 4-hour pot life once the hardener goes in, cannot be rehardened, one-year unopened shelf life against three years for both Varathane products. If you finish four projects a year, that's a $153 gallon with a four-hour clock on it and a twelve-month expiration date. Its one real perk is that Bona doesn't require abrading between waterborne coats unless more than 48 hours have passed.
There's also a regulatory wrinkle. EPA's national architectural coatings rule dates to 1998, amended in 2000, at 40 CFR Part 59 Subpart D, but states set tighter numbers. Virginia caps varnishes at 350 g/L and floor coatings at 250. Minwax's oil sheet lists 450 g/L maximum and Varathane's oil floor finish 275. That's why the same brand's oil product gets reformulated or disappears depending on where you live. Minwax even says on its own sheet that compliance drove higher resin concentrations and a thicker consistency.
Settle It on Your Own Offcuts
No source I have settles the durability question, and I'm not going to pretend otherwise. What Flexner gave readers in 2018 is better than a verdict anyway: a method. Push architect's drawing pencils from 2B up through 5H across the cured finish until one cuts in, and the previous pencil is your rating. That's the shop version of ASTM D3363, the same test behind Varathane's published 3H. Puddle water under an upturned cup so it can't evaporate and check every ten minutes. Set a hot metal container of water on it, then repeat with wet cheesecloth underneath. Drop mustard, vinegar, juice and alcohol on it and sponge them off at intervals. Crosshatch it with a razor, press masking tape down, rip it off fast, look at the edges.
Finish two offcuts off the same board, one each way, wait out the cure, and wreck them both. That'll tell you more about your wood, your shop and your humidity than any data sheet, and it costs you a scrap and an afternoon. If you're not willing to do that, apply the third coat and stop arguing.