Nail and Screw Sizes: Penny Sizes, Gauges, and What the Code Actually Asks For

The chart everyone publishes stops one column too early

Search for nail sizes and you get the penny table: 8d is two and a half inches, 16d is three and a half, and so on down the list. It is correct and it is the same on every site.

It is also only half of a nail. Length does not carry load. Diameter does. And the penny size, which is purely a length designation, says nothing about it.

You can see how much this matters by looking at what the building code writes down. IRC Table R602.3(1), the fastening schedule, never asks for "3-16d". It asks for:

3-16d common (3-1/2″ × 0.162″)

Every single entry carries both numbers. The code does that because the penny size on its own does not identify a nail, and the schedule is not going to leave that to chance.

And it goes further than that. Look down almost any row of the schedule and the last option offered is something like:

or 4-3″ × 0.131″ nails

No penny size at all. Just a length and a diameter. That is the pneumatic nail — the one that comes in a coil or a strip and goes into a framing gun, and the one most framing in America is actually built with. The code identifies it the only way it can, which is the way it would rather identify all of them.

A 16d common and a 16d box are not the same nail

Same name. Same length. Different fastener.

• 16d common: 3-1/2 inches long, 0.162 inch thick
• 16d box: 3-1/2 inches long, 0.135 inch thick

That is twenty percent more diameter on the common — and because strength follows cross-sectional area rather than width, forty-four percent more steel.

A connection the schedule specifies as 16d common is not satisfied by a 16d box, and the box on the shelf that just says "16D" is not telling you which you have. Look for the diameter on the label, or measure one with calipers.

The code's substitutions are doing arithmetic, not being generous

Here is the part that makes the whole schedule make sense.

Read one line properly — ceiling joists to top plate:

4-8d box (2-1/2″ × 0.113″); or 3-8d common (2-1/2″ × 0.131″); or 3-10d box (3″ × 0.128″); or 3-3″ × 0.131″ nails

Four options, and the counts are different. It looks like the code is offering flexibility. Work out the total cross-section of steel each option puts through the joint:

• 4 × 0.113 inch nails → 0.0401 sq in
• 3 × 0.131 inch nails → 0.0404 sq in
• 3 × 0.128 inch nails → 0.0386 sq in
• 3 × 0.131 inch gun nails → 0.0404 sq in — the same shank as the 8d common, just longer

Three of the four are identical to within one percent, and the spread across all four is under five. The schedule is trading diameter for count and holding the steel constant.

Which tells you the thing worth knowing: you cannot invent your own substitution. Three box nails where the code says four is not "close enough" — it is a quarter of the connection missing, and the schedule already did that sum for you.

The penny system works, then stops working

From 2d to 10d there is a genuine rule:

length = 0.25 × penny + 0.5 inches

So 2d is 1 inch, 6d is 2 inches, 10d is 3 inches, and every step is a clean quarter inch. It is the kind of pattern that makes people think they have understood the system.

Then it breaks. The formula puts a 16d at four and a half inches. A 16d is three and a half. From 12d up, the penny sizes are a list you memorise rather than a rule you apply — 12d is 3-1/4, 16d is 3-1/2, 20d jumps to 4.

Screw gauges are a position on a scale

The gauge number on a wood screw is not inches, not millimetres, and not a count of anything. It is an index into a scale that begins at #0 = 0.060 inch and adds thirteen thousandths per step:

diameter = 0.060 + 0.013 × gauge

A #8 is 0.060 + 0.104 = 0.164 inch. A #12 is 0.216. The formula is exact, not a rule of thumb, and it holds for every gauge in the chart above.

Two things fall out of it. The steps are small in absolute terms — a #8 and a #10 are twenty-six thousandths of an inch apart — but they are not small in what matters: withdrawal strength climbs with diameter, so going from a #8 to a #10 is a sixteen percent bigger screw, not a rounding.

Pilot holes, and the clearance hole nobody drills

A pilot hole is sized to clear the root of the screw, not the thread. The thread still needs wood to cut into; what you are removing is the solid core the screw cannot compress out of the way.

In softwood that means a bit around two-thirds of the major diameter. In hardwood, go one bit larger — hardwood does not compress, it splits, and a pilot that works in pine will crack oak.

And then the part that gets skipped. If you are pulling two boards together, the top board needs a clearance hole at the full major diameter, all the way through. Not a pilot — a hole the screw passes through without touching. Otherwise the thread bites the top board as well as the bottom one and cheerfully holds the gap open while you tighten. Anyone who has ever driven a screw and watched the joint refuse to close has met this.

One honest caveat on the pilot chart above: unlike the nail diameters, which come from the code, pilot sizes are trade convention rather than a standard. They shift with species, moisture and screw design, and modern self-drilling deck screws often want no pilot at all in softwood. Start there, then test in an offcut of the wood actually in front of you.

Where the schedule changes

The fastening schedule is one of the most amended parts of the code. High-wind regions and seismic zones require more fasteners, larger fasteners, or engineered connectors instead — and the numbers here are the baseline the amendments start from, not the last word.

If you are somewhere the wind matters, the schedule in your adopted code is the one to read, and there will be more to it than this page.