How to cut a sheet of plywood with the least waste

A full sheet of plywood is 2440 × 1220 mm, or 8 × 4 feet. That is 2.98 square metres of material you have already paid for. How much of it ends up in the finished piece is decided long before you switch the saw on.

Waste comes from three places, and only one of them is obvious.

The blade eats a strip out of every cut

Every cut turns a few millimetres of plywood into dust. That width is the kerf, and on a track saw or a table saw it is usually around 3 mm — an eighth of an inch.

Three millimetres sounds like nothing. Do the arithmetic across a real job.

Suppose you need six shelves, each 400 mm wide, from one sheet. Six cuts at 3 mm is 18 mm of dust. 1220 divided by 400 gives three shelves per width with 20 mm left over — but only if the blade is infinitely thin. In reality the first cut costs you 3 mm, the second another 3 mm, and the 20 mm of slack is gone before you reach the third shelf. You get two shelves and an offcut instead of three.

This is the single most common way a cutting plan drawn on paper fails at the saw. If you plan by dividing the sheet dimensions by the part dimensions, you have planned for a blade with no thickness.

What to do: measure your actual kerf once. Cut a test piece, measure the offcut and the parent, and subtract. Write the number on the saw in marker. Then include it in every calculation — one kerf for every cut, not one for every part.

The grain runs one way and it halves your options

On veneered plywood the face grain runs the length of the sheet. For a cabinet side, a door or anything else that will be seen, the grain has to run the way you want it to run in the finished piece — usually vertically.

That single constraint removes half of the packing options. A 600 × 400 part that must follow the grain cannot be turned to 400 × 600 to fill a gap, no matter how neatly it would fit.

The parts that do not care — a cabinet back, a drawer bottom, anything painted or hidden — should be free to rotate. Fixing the grain on every part because it matters on some of them is a quiet, expensive habit.

What to do: mark each part on your list as grain-locked or free, one at a time, before you plan anything. It is the highest-leverage two minutes in the whole job.

Every cut has to go all the way across

A track saw, a circular saw with a guide, and a panel saw all do the same thing: they cut in a straight line from one edge of the piece to the other. You cannot stop halfway and turn a corner.

This is called a guillotine constraint, and it is why a layout that looks efficient can be impossible to cut. If your plan has a small part tucked into the middle of a sheet with three different parts around it, there is no sequence of edge-to-edge cuts that frees it.

A layout is only real if it can be reduced to a sequence: cut the sheet in two, set one half aside, cut the other in two, and so on until every piece is a finished part or an offcut.

What to do: before you commit, check your plan by working backwards. Take any part and ask which single edge-to-edge cut separated it from its neighbour. If you cannot answer for every part, the plan will not survive contact with the saw.

Cut the big pieces first, and in the right direction

Two practical rules that cost nothing:

Break the sheet down before you chase precision. A full sheet is heavy, floppy, and awkward to hold against a fence. Make one or two rough cuts to get pieces you can handle, leaving 10–20 mm of extra material, then cut those to final size. Accuracy on a piece you can actually control is far better than accuracy you fight for on a wobbling 2440 mm sheet.

Rip along the length first when your parts are long. If most parts share a width, ripping the sheet into strips of that width first means every subsequent cut is a short crosscut. Fewer long cuts means fewer chances for the blade to drift.

Keep the waste in one piece

Two plans can waste exactly the same total area and be worth completely different amounts to you.

Plan A leaves eleven strips 40 mm wide scattered across two sheets. Plan B leaves one rectangle 700 × 500 mm. Both waste 0.35 m². Plan A gives you firewood. Plan B gives you the sides of the next drawer.

This is the thing a naïve “minimise total waste” calculation gets wrong, and the thing an experienced hand does without thinking: push the leftovers together. When two layouts are level on waste, take the one whose offcut is rectangular and big.

What to do: when you compare plans, do not just read the waste percentage. Look at the shape of what is left.

A worked example

Say you are building a simple cabinet: two sides at 800 × 400, one top and one bottom at 800 × 400, a back at 780 × 380, and four shelves at 780 × 380. All from one 2440 × 1220 sheet of 18 mm birch ply, 3 mm kerf, sides and top grain-locked along the length.

Done badly — parts placed in list order, grain locked on everything, no account of the kerf — this needs two sheets, and the second one is barely touched.

Done well it fits on one, because:

  • the back and the shelves are hidden, so they rotate freely and slot into the strip left over from the sides;
  • the kerf is subtracted before the third part goes into each strip, so no cut is planned into space that the blade has already eaten;
  • the leftover is consolidated at one end of the sheet, roughly 1220 × 500 — a usable piece rather than four strips.

The difference between the two plans is one sheet of birch ply. At the price of 18 mm birch, that is a meaningful amount of money for a job this small, and it is entirely decided at the planning stage.

Where software comes in

None of the above needs an app. A pencil, a tape measure and twenty minutes will get you a good plan for a simple job, and doing it by hand a few times teaches you more than any tool will.

What software is good for is the part that is genuinely tedious: trying hundreds of arrangements of the same parts and keeping the best one, while never forgetting the kerf on any of them. That gets valuable as soon as you have more than about a dozen parts, or when you change one dimension and have to redo the whole thing.

SawKit does exactly that job on the Mac: you type the parts, it works out the layout across as many sheets as it takes, tells you the order to make the cuts in, and prints the plan for the shop. It is bought once rather than rented, and it works with no signal — which matters, because workshops are usually the worst room in the building for a connection.

If you want to see a layout for your own parts first, the plywood cut calculator here is free, needs nothing installed, and counts the kerf on every cut.