Sandpaper Grit Chart: CAMI and P-Grade Equivalents, and Where to Stop on Bare Wood
Up to 220, CAMI and FEPA P-grades track each other closely enough to swap freely. Past that they diverge hard — a CAMI 400 is roughly a P800, not a P400. On bare wood most work should stop at 180 for dense closed-pore species and 220 for open-pore ones, because going finer burnishes the surface rather than smoothing it.
Below 220 the two numbering systems are close enough to treat as identical: a CAMI 120 and a P120 remove roughly the same amount and leave roughly the same scratch. Above 220 they part company fast. CAMI 400 sits near P800. CAMI 600 sits near P1200.
On bare wood, most projects should stop at 180 or 220 regardless of system. Past that you are polishing rather than cutting, and a burnished surface takes finish unevenly — the classic cause of blotchy cherry and maple that gets blamed on the finish rather than the prep that came before it.
This page covers bare-wood surface preparation only: milling marks through to the last dry pass before anything goes on the wood. What happens after that is somebody else's page.
CAMI to P-grade, side by side
CAMI is the US standard, printed as a plain number. FEPA is the European one, printed with a P prefix. The micron figures below are average particle size, and the reason the systems diverge is that FEPA screens far more tightly at the fine end — a P400 sheet has a much narrower spread of particle sizes than a CAMI 400, so it cuts more uniformly and leaves a more consistent scratch pattern.
CAMI grit FEPA equivalent Average particle size, microns (CAMI / P) What it is genuinely for 40 P40 425 / 425 Heavy removal on a belt sander. Not a hand grit 60 P60 265 / 269 Levelling glue squeeze-out and severe machine burn; leaves scratches that take two grits to clear 80 P80 190 / 201 The honest starting grit for planer snipe and jointer chatter 100 P100 140 / 162 First grit that leaves a surface worth calling flat 120 P120 115 / 125 Standard start on stock that came off the machine clean 150 P150 92 / 100 Where softwood can stop 180 P180 78 / 82 Where dense closed-pore hardwood should stop 220 P220 66 / 68 The practical end of bare-wood prep for most work 240 P280 53 / 52 Systems start diverging here — a CAMI 240 is a P280 320 P400 36 / 35 Rarely useful on bare wood; useful on end grain 400 P800 23 / 22 A CAMI 400 is nowhere near a P400 600 P1200 16 / 15 Turned work and burnished surfaces only 800 P1500 12 / 13 1000 P2000 9 / 10 Use the middle column to swap between systems mid-progression without accidentally skipping three grades. Anything below 220 is safe to substitute freely; above it, read the row rather than the number. Where the two systems stop agreeing, and why it bites
The failure looks like this. You start a progression with a CAMI-graded belt at 120, move to a P180 disc because that is what was in the drawer, then finish with CAMI 220. Fine — those three are effectively 120, 180, 220.
Now try the same thing at the top of the range. You finish at CAMI 320 thinking you have gone one step past 220, when you have actually gone to P400 and skipped nothing in particular. Or worse, you reach for P400 believing it follows CAMI 320, and it is barely a step at all.
A second wrinkle: the same grit number on different backings cuts differently. A 180 grit on a stiff paper backing on a random orbit sander cuts far more aggressively than 180 on a flexible sponge block, because the paper holds the abrasive proud while the sponge lets it conform and glide. Grit number describes particle size, not aggression.
The progression: how far you can jump
Each grit exists to remove the scratches left by the grit before it. Skip too far and the new abrasive cannot reach the bottom of the old scratches, so you polish the peaks and leave the valleys — invisible on dry bare wood, glaringly obvious the moment anything wets the surface.
Current grit Safe next step Aggressive next step Cost of the aggressive jump Verdict 80 100 120 Roughly double the time at 120 Take the jump. 100 rarely earns its place 100 120 150 About 50 percent more time at 150 Fine on softwood, marginal on oak 120 150 180 50-70 percent more time at 180 The most common skip, and usually justified 150 180 220 Doubles time at 220 and often fails on dense stock Do not skip here on hardwood 180 220 320 Will not clear 180 scratches at all Never. This jump is a wasted sheet 220 320 400 Only relevant on turned or end-grain work Stop at 220 unless you have a specific reason Cross-reference your starting grit with where you want to end up. The rule underneath: you can skip one step, never two, and skipping always costs time at the next grit rather than saving it overall. There is a cheap test for whether a grit has done its job. Wipe the surface with a rag dampened with water or mineral spirits and look across it at a low angle against a light. Every scratch the last grit failed to remove shows up immediately. Let it dry fully, knock back the raised fibres, and carry on — that is not a finishing step, it is an inspection step, and it is the single most useful habit in surface prep.
Stop grit by species
How fine to go is a property of the wood, not of ambition. Dense, tight, diffuse-porous species burnish — the abrasive stops cutting and starts polishing, closing the surface so it no longer absorbs evenly. Open ring-porous species keep taking abrasive to a higher number because the pores are still there whatever you do to the flats.
Species Janka hardness (lbf) Pore structure Stop grit, penetrating finish Stop grit, film-forming finish Why Eastern white pine 380 Closed, soft 150 180 Softwood compresses under the disc; finer grits just polish the earlywood Poplar 540 Diffuse porous 150 180 Fuzzes readily above 180 rather than getting smoother Cherry 950 Diffuse porous, tight 180 220 Burnishes easily, and blotches worse the finer you go Soft maple 950 Diffuse porous 180 220 Same failure mode as cherry, slightly more forgiving Black walnut 1,010 Semi-ring porous 180 220 Open enough that 220 still leaves it absorbent Ash 1,320 Ring porous 180 220 Pores dominate the look; grit choice only affects the flats Red oak 1,290 Ring porous, very open 180 220 You will never close these pores by sanding, so do not try White oak 1,360 Ring porous, tyloses-filled 180 220 Tighter than red oak but still open enough for 220 Hard maple 1,450 Diffuse porous, very tight 180 180 Above 180 it stops absorbing. This is the classic blotch cause Hickory 1,820 Ring porous, hard 150 180 Hard enough that fine grits mostly burnish rather than cut Pick your species and finish class and stop there. If the last column says burnishing is the risk, going one grit finer makes the result worse, not better. End grain is the exception to all of it. End grain absorbs far more than face grain, so taking it two grits finer than the faces — 320 where the face stopped at 220 — closes it enough to even out. That is the one place on a bare-wood surface where the extra step genuinely pays.
Milling marks are not a sanding problem
If a board comes off the planer with visible ridges, the honest fix is upstream. Sanding out planer marks at 80 grit costs half an hour per face and removes enough material to matter on thin stock.
Ridges spaced evenly across the width are a nicked knife. Rotate or index the knives rather than sanding. Scalloping down the length is normal knife-mark spacing and gets shorter with a slower feed rate on machines that have one. Snipe at the ends is a pressure roller or outfeed support issue.
Every minute spent on machine setup buys back roughly ten at the sander. That trade is not close.
Machine sanding and hand sanding do different jobs
A random orbit sander is a levelling tool. It does not know or care about grain direction, and its scratch pattern is a swirl of overlapping arcs. Run it flat, let its own weight do the work, and move it roughly one inch per second — leaning on it stalls the orbit and turns a random pattern into a circular one, which is exactly the mark that shows up as pigtails later.
Hand sanding with a block along the grain is a finishing move, not a levelling one. One pass at the final grit, with the grain, removes the sander's swirl and leaves a scratch pattern the eye reads as grain rather than as damage.
A drum or wide-belt sander is neither — it is a thicknessing machine that happens to use abrasive, and the marks it leaves run straight down the length. Those need the same final hand pass.
Dust extraction matters more than most people allow. An orbital sander running without collection re-cuts its own swept dust, which loads the disc, drops the effective cut rate and puts the finest particles you will breathe all week into the air at head height. Collection at the pad is not a nicety on this machine.
Raising the grain, and the last dry pass
Sanding lays fibres over rather than cutting them all off. Wet the surface, those fibres stand back up, and the previously smooth board turns rough.
Do it deliberately instead of being ambushed. After the final grit, wipe the surface evenly with a damp rag, let it dry for thirty minutes, then knock the raised fibres back with the same grit you finished on — very lightly, one pass. The second raising is far smaller than the first, and on most work one cycle is enough.
Skip this only if nothing water-based is going near the wood. Then vacuum rather than blow, because compressed air drives dust into open pores where it sits waiting. A vacuum with a brush head followed by a wipe with a clean cloth is the last thing that should touch bare wood on this site — what happens after that is outside what we cover here.