Every old house has one piece of trim that somebody painted eleven times. Door casings, a mantel, the run of baseboard in the front hall—somewhere there's a profile underneath all those layers, and the temptation is to come at it with the coarsest thing in the toolbox. That's how you end up with a rounded, mushy shadow of the original molding. A rotary tool is the right instrument here precisely because it's small and slow, but only if you treat the profile as the thing you're preserving and the paint as the thing you're removing.
What you will learn
- When stripping is worth it and when a new board is smarter
- Why the flats go fast and the coves eat your weekend
- Which abrasive shapes match which parts of a profile
- How to tell paint layers apart and what lead means for your method
- What to do before the primer so the profile reads crisply
Is this piece even worth stripping?
Sometimes the honest answer is no. Milling a new length of baseboard is cheap, and if the profile is a simple ogee that's still in catalogues, replacement beats restoration on time alone. Strip it when the profile is unusual, when the run is continuous and matching new stock would be obvious, or when the wood itself is old-growth and denser than anything you can buy. Start by scraping a small test patch down to bare wood in an inconspicuous spot. If the wood underneath is grey and soft, you're looking at rot and the job just changed. If it's clean and pale, proceed. For the removal itself, having every abrasive shape available is the difference between a crisp result and a rounded one—the Orange Black Ultimate 155-piece corded system covers the drum sanders, flap wheels, and detail bits in one case.
Why the flats go fast and the coves don't
Because a flat surface lets any abrasive do its job, and a concave one demands an abrasive shaped like the cove. Most of a molding's visual weight is in the curves, and those are exactly the places a flat sanding block can't reach. This is where people give up and reach for chemical strippers on the whole thing. Chemical stripper is fine on the flats and mediocre in the coves because it softens paint into a slurry that pools in the low spots and then sets again. My approach is the reverse of the usual order: strip mechanically where the shape is complex, chemically where it's flat. The rotary does the coves, the flats, and the beads; the stripper does the broad faces.
Which abrasive shape goes where?
Match the tool to the geometry. Flap wheels and small drum sanders for the flats and the gentle convex rolls. A pointed detail bit for the tight coves and the corners where two curves meet. Felt or rubber-impregnated points for the very small beads, because they flex enough to follow the curve without flattening it. Change the abrasive as soon as it loads—a loaded flap wheel stops cutting and starts burning, and burnishing paint into the wood is worse than leaving it. Start coarser than feels right on the first pass to break through the top layers, then step down. For fine bead work, a lighter tool helps because you're making very small movements and a heavy body fights you. The Essential White 70-piece set is what I keep on the bench for the detail passes.
How do you read the paint layers?
Sand a small spot through to wood on a bevel, then look at it straight-on in good light. Each color band is one generation of paint, and the count tells you the age and roughly how many different application methods you're fighting. Oil paint goes to powder and comes off easily; modern latex softens into a rubbery film that clogs abrasives, and that's the stuff that makes people hate this job. For latex, a heat gun on low with a putty knife underneath works better than sanding because you're lifting softened sheets rather than turning them into dust. Then finish the residue with abrasives. Nothing about this step is fast. Budget more time than you think and do it in sessions rather than one long push—fatigue is when you round off the profile.
What if it's lead paint?
Then stop and change the method. Dry sanding lead paint is the worst thing you can do in a house. Wet the surface, work by hand or with a low-speed tool that doesn't throw dust, use a HEPA vac for cleanup, and keep the area sealed off from the rest of the house. Test kits are inexpensive and worth running on anything in a pre-1978 property before a single abrasive touches it. If it's positive and the run is long, encapsulation or professional removal may genuinely be the better call. I'm not going to tell you to power through it—the cleanup is the expensive part, and it's invisible until someone swabs a surface.
How do you keep the profile crisp?
Two habits. First, never let an abrasive dwell on an area. Edges are thin and heat up faster than flats, and a rounded aris is what makes restored trim look restored by amateur. Keep moving. Second, check your work by feel and not just by eye—run a finger along the profile and you'll find flat spots that look fine from two feet away. Where two curves meet, the junction should still be a distinct line. If it's soft, you've crossed the line between removing paint and removing wood. Where the run has repeating units—dentil blocks, for instance—do one, get it right, then match the rest to it rather than doing them all at your current skill level and discovering the drift at the end.
What happens between bare wood and primer?
A clean pass at 220, then a wipe with mineral spirits to show you what's left. Mineral spirits reveal remaining paint ghosts and old filler that look identical to bare wood when dry. Fix those now, because primer will not hide a shadow and the topcoat will make it worse. Fill any nail holes and open joints, let the filler cure fully, then sand the filler flush—filler that looks flush when soft always sinks slightly and leaves a dish. On a piece this detailed, a 6-inch metal rechargeable caliper is surprisingly useful for checking that a replacement section matches the original profile depth before you commit to milling it.
What's a realistic timeline?
A single door casing with a moderate profile is a solid evening of work, three to four hours, and that's after any chemical stripping has done its dwell time. A full mantle with returns and a shelf edge is a weekend. Most of that time is in the coves and beads, not the flats, which is why quoting by linear meter is a good way to lose money.
From the workshop: the mantel I rounded off.
A 1920s mantle, eleven layers, and I did the whole thing with one flap wheel because it was the abrasive already on the tool. I got through the flats beautifully and destroyed the two narrow beads flanking the center panel—by the time I noticed, the beads were gone and the shelf nose had a soft roundover where there had been a crisp aris. It read as "old trim that's been sanded a lot," which is the tell. I milled a replacement bead profile on a router table and spliced it in, which took another six hours and still isn't invisible from three feet. On the next one—the front hall baseboard—I used four different abrasive shapes and stopped to check the junction lines every meter. That run looks original. Same house, same paint, about four hours' difference in total effort.
Common mistakes recap
- One abrasive shape for the whole profile
- Dwelling on arrivals and rounding them off
- Dry sanding before testing for lead
- Skipping the mineral spirits wipe and discovering ghosts after primer
- Sanding filler before it has fully cured
- Trying to finish a long run in one session
FAQ
Should I use chemical stripper instead?
Both, in sequence. Stripper for the broad flats where it works well, and rotary for the coves and beads where slurry pools and fails. Using only one of them is why most people give up halfway.
How do I know when I've hit bare wood?
Colour and dust. Paint dust is tinted and fine; wood dust is pale and fibrous. When the dust changes color, you're through, and that's the moment to stop on that spot.
Can I strip trim without taking it off?
Yes, and it's usually better. Removing old trim breaks it and the wall behind it. Mask the wall and floor, work in sections, and accept that your neck will complain.
What grit should I finish at?
220 for paint, 320 if you're staining rather than painting. Going finer than that closes the grain and can make stain uptake blotchy on old wood.
Why does the paint come off in sheets in some spots and powder in others?
Different generations of paint. Older oil-based layers powder; newer latex softens. That variation is normal, and it's why no single method handles a whole piece.
Is a heat gun safe on old trim?
On low, moving constantly, yes. It's the fastest way through latex. Hold it too close and you scorch the wood, and on lead paint it's a no—you don't want to vaporize anything.
How do I match a replacement profile?
Measure the depth and width of each step in the profile with a caliper and transfer those numbers to the new stock. Eyeballing a profile is how you end up with a splice everyone notices.
Do I need to seal bare old wood before priming?
A stain-blocking primer is worth it on old trim, especially where water staining or smoke has soaked in. It saves a second coat of finish later.
