Frame-and-panel cabinetry showing floating solid panels within stile-and-rail frames

Cabinet Construction & Joinery

Wood Movement

Wood barely changes length and changes width constantly. Floating panels, quartersawn stock and panel clearance all follow from that one asymmetry.

10 min readUpdated July 2026Reviewed by Andrey Voronenko

Quick Answer

Why does wood cabinetry move with the seasons?

Wood is hygroscopic: it exchanges moisture with the air until the two reach equilibrium, then changes dimension as that equilibrium shifts. The change is wildly uneven by direction. The Forest Products Laboratory puts average shrinkage along the grain at just 0.1 to 0.2 percent from green to oven-dry, while across the grain it is one to two orders of magnitude larger.

Everything a well-built cabinet does follows from that asymmetry. Narrow parts with their length along the grain — stiles, rails, face frames — hold the shape. Wide parts that must change width are left free to do it, which is what a floating panel in a grooved frame is.

Key Takeaways

  • Wood moves 10 to 100 times more across the grain than along it. Longitudinal shrinkage averages 0.1 to 0.2 percent green to oven-dry; tangential is twice radial.
  • The equilibrium moisture content curve is not linear. At 70 degrees Fahrenheit, 30 to 50 percent relative humidity adds three points of moisture content; 70 to 90 percent adds more than seven.
  • Below about 30 percent moisture content — the fibre saturation point — every moisture change changes size. Above it, wood is stable. Cabinetry lives permanently below it.
  • What quartersawing buys is the tangential-to-radial ratio, which varies by species: roughly 2.3 to 1 in American beech and red oak, but 1.3 to 1 in yellow birch.
  • Two Californian cities can have opposite seasonal cycles. Fresno dries through the summer; Los Angeles moistens through it, over a third of the amplitude.

Wood Chases the Room, Slowly and Forever

Wood is hygroscopic: it exchanges moisture with the air until the two are in equilibrium, and then the air changes.

The Forest Products Laboratory calls the resulting figure equilibrium moisture content. At 70 degrees Fahrenheit, wood sits at about 4.5 percent moisture content in 20 percent relative humidity, 6.2 percent at 30 percent, 9.2 percent at 50 percent, 13.1 percent at 70 percent, and 16.0 percent at 80 percent. That row is not a straight line: 30 to 50 percent relative humidity adds three points, while 70 to 90 percent adds more than seven. The top of the range is disproportionately dangerous, which is why a bathroom punishes cabinetry in a way a dry winter never does.

Two caveats belong with those numbers. The tabulated values were derived primarily for Sitka spruce under oscillating vapour pressure desorption, chosen because it sits midway between adsorption and desorption — a practical compromise rather than a species-specific reading. And sorption hysteresis means wood arriving at a given humidity from the wet side settles higher than wood arriving from the dry side, at a ratio averaging about 0.8 near room temperature.

Above the fibre saturation point — which averages about 30 percent moisture content but varies by several points between species and pieces — wood stops changing dimension altogether. Below it, every change in moisture content is a change in size. Cabinetry lives permanently below fibre saturation, which is another way of saying a cabinet is never dimensionally at rest.

Three Directions, Three Different Answers

The degree of anisotropy in wood is the most useful fact in cabinetmaking — not a modest difference between directions but one to two orders of magnitude.

Along the grain, wood barely moves: average longitudinal shrinkage from green to oven-dry runs between 0.1 and 0.2 percent for most species. Across the grain it shrinks most in the direction of the annual rings — tangentially — and about half as much across those rings, radially. A door stile, cut with its length along the grain, is close to dimensionally inert. A wide panel is not.

Two exceptions matter enough that the Wood Handbook names them. Reaction wood — compression wood in softwoods, tension wood in hardwoods — and juvenile wood from near the centre of the tree can shrink 2 percent lengthwise green to oven-dry, ten to twenty times the normal figure. Because it can occur in the same board as normal wood, the result is not uniform shortening but bow, crook, twist, and cross breaks — which is why a rail that leaves the mill straight can be a spring by the time the door is hung.

For stock already dried to service moisture content, the useful figures are the dimensional change coefficients published for the 6 to 14 percent range, expressing change per 1 percent change in moisture content. The arithmetic is explicit: starting dimension times coefficient times change in moisture content. The Wood Handbook example takes a flat-grained white fir board 9.15 inches wide at 8 percent moisture content and finds it 0.067 inch wider at 11 percent. One caveat travels with it: the variability of transverse and volumetric shrinkage runs to a coefficient of variation of approximately 15 percent, and predicting the shrinkage of an individual piece is, the Wood Handbook says plainly, impossible.

What Quartersawing Actually Buys, by Species

The argument for quartersawn stock is usually made on appearance. The structural argument is stronger — but worth far more in some species than others.

A quartersawn board is oriented so that seasonal width change follows the radial coefficient rather than the tangential one, so the prize is the ratio between the two — and that ratio varies more between species than most specifications acknowledge. In American beech and commercial red oak it is roughly 2.3 to 1, so quartering removes well over half the movement; in eastern white pine about 3 to 1; in yellow birch only about 1.3 to 1.

The two directions combined are also what distorts a board rather than merely resizing it. Because tangential shrinkage is about twice radial and the rings are curved, the Wood Handbook illustrates flat-sawn boards cupping away from the heart, squares becoming diamonds, and rounds becoming ovals. A quartersawn board carries that curvature through its thickness instead of its width, so it stays flatter as well as narrower — which is why the Architectural Woodwork Standards require the method of slicing to be specified for wood casework rather than left to the shop.

Dimensional change coefficients from the Forest Products Laboratory Wood Handbook, expressing change per 1 percent change in moisture content within the 6 to 14 percent range. The ratio column is the tangential coefficient divided by the radial one, which is precisely what quartersawing converts.
SpeciesRadial coefficientTangential coefficientTangential-to-radial ratioWhat quartersawing is worth here
American beech0.001900.004312.3The largest payoff here. Flat-sawn beech is among the most active woods in the tables; quartered, it is ordinary.
Red oak (commercial)0.001580.003692.3A wider ratio than white oak, so the structural case for quartering is at least as strong. The ray fleck is why people do it.
Sugar maple0.001650.003532.1Quartering roughly halves seasonal width change. Maple reads as a calm material and is not one dimensionally.
True hickory0.002590.004111.6High in both directions. Quartering helps, but hickory moves however it is cut.
White ash0.001690.002741.6Moderate movement and ratio. Forgiving in wide components without being quartered.
Yellow birch0.002560.003381.3The exception. A high radial coefficient means quartering buys little — the two directions are already close.
Eastern white pine0.000710.002123.0The widest ratio here. Quartersawn it is remarkably stable; flat-sawn, three times more active.
Dimensional change coefficients from the Forest Products Laboratory Wood Handbook, expressing change per 1 percent change in moisture content within the 6 to 14 percent range. The ratio column is the tangential coefficient divided by the radial one, which is precisely what quartersawing converts.

Why Panels Float, and What the Standards Do When They Cannot

The frame-and-panel door has never been improved on, because the problem it solves has never gone away.

Stiles and rails are narrow with their length along the grain, so they hold the door stable. The panel is wide and must change width seasonally, so it sits in a groove with clearance at its edges, and where glue is used at all it goes at the centre of the top and bottom rails so movement is shared evenly. The Forest Products Laboratory gives essentially this instruction: design and install large solid pieces so panels are free to move across the grain, and prefer narrow widths.

When a panel is too wide for solid wood to be sensible, the Architectural Woodwork Standards stop using solid wood. They permit solid lumber raised panels when the width does not exceed the Custom Grade limit, and require a rim raised panel — membrane-wrapped veneer over a stable panel product — for Premium Grade, whenever widths exceed that limit, and whenever a transparent finish is used. Past a certain width, the honest answer to wood movement is a substrate that does not move.

The same instinct runs through the joinery. The Wood Handbook warns that adhesive bonds are damaged when adjacent pieces have different shrinkage coefficients — different species, or radial grain bonded to tangential — and recommends bonding at the moisture content the assembly will actually see.

What a Season Actually Does, in California Terms

National guidance flattens something that matters in a state this large: two Californian cities can have opposite seasonal cycles.

The Forest Products Laboratory tabulates equilibrium moisture content for outdoor conditions in United States cities, averaged from thirty or more years of temperature and humidity data. Fresno runs from 16.4 percent in January down to 7.8 percent in July — a swing of nearly nine points, with the dry extreme in high summer. Los Angeles runs the other way and far more gently: 12.2 percent in January, 15.0 percent in July, peaking at 15.1 in August, a total range of about three points. Conditioned interiors compress those outdoor figures, but the shape of the two curves survives, and the shape is the design input.

For installation moisture content the Forest Products Laboratory recommends an average of 8 percent for interior woodwork across most of the United States, 6 percent in the dry southwestern area, and 11 percent in the damp warm coastal area, with individual pieces held to 6 to 10, 4 to 9, and 8 to 13 percent respectively. California spans a wide band of that guidance — the reason a shop building statewide asks where a kitchen is before deciding how wide a solid panel may be.

Shrinkage begins at the fibre saturation point and proceeds fairly linearly as wood dries, so a cabinet responds to the size of the swing rather than the position of the midpoint. A house held reliably at 45 percent relative humidity and one held reliably at 35 percent will both be fine. One that alternates between them every few weeks opens joints.

Frequently Asked Questions

How much does a cabinet door panel actually move in a year?
It depends on width, species, cut, and the size of the humidity swing. The Forest Products Laboratory publishes the arithmetic — starting width times the dimensional change coefficient times the change in moisture content — and its example takes a flat-grained white fir board 9.15 inches wide at 8 percent moisture content and finds it 0.067 inch wider at 11 percent.
What is equilibrium moisture content, and what number should I aim for?
It is the moisture content wood settles at for a given temperature and humidity. At 70 degrees Fahrenheit the Wood Handbook puts wood at about 6.2 percent in 30 percent relative humidity, 9.2 percent at 50 percent, and 13.1 percent at 70 percent. The Forest Products Laboratory recommends installing interior woodwork at an average of 8 percent across most of the United States.
Why does wood move across the grain but not along it?
Because the cell walls that swell are aligned with the grain. The Forest Products Laboratory puts average longitudinal shrinkage from green to oven-dry at just 0.1 to 0.2 percent, while across the grain the figures are one to two orders of magnitude larger, with tangential about twice radial. That is why a door stile is almost inert and a wide panel in the same door is not.
Is quartersawn wood always worth the premium for stability?
Not equally, because the benefit is exactly the ratio between the tangential and radial coefficients. Using the Forest Products Laboratory figures, that ratio is roughly 2.3 to 1 in American beech and commercial red oak and about 3 to 1 in eastern white pine — but only about 1.3 to 1 in yellow birch, so quartering birch for stability alone buys little.
What is the fibre saturation point?
The moisture content at which the cell walls are fully saturated but no liquid water sits in the cavities. The Forest Products Laboratory notes it averages about 30 percent, varying by several points between species. Above it wood does not change dimension; below it, every moisture change changes size — and cabinetry lives permanently below it.
Why does a door panel show an unfinished line at its edge?
Because the panel was finished at a wider dimension than it now occupies. A floating panel is finished in place, so the coating stops at the edge as it then stood; when the panel shrinks it withdraws from the groove and exposes wood the finish never reached. It is evidence the panel is floating correctly — one glued rigidly into its frame would have cracked instead.
Does the same cabinet behave differently in different parts of California?
Yes, and in opposite directions. The Forest Products Laboratory tabulates outdoor equilibrium moisture content by city from thirty or more years of data: Fresno runs 16.4 percent in January down to 7.8 percent in July, while Los Angeles runs 12.2 percent in January up to 15.0 percent in July, a total range of about three points. Those are outdoor figures, but the direction and relative amplitude carry through.
Can wood movement be predicted precisely for a specific board?
No, and the Wood Handbook says so directly: predicting the shrinkage of an individual piece is impossible, while the average of a quantity can be predicted accurately. It puts the variability of transverse and volumetric shrinkage at a coefficient of variation of approximately 15 percent. Published coefficients are design allowances, which is why good detailing gives a panel more clearance than the arithmetic requires.

Explore

Movement Is Coupled to Everything Else

What a panel does seasonally decides what its joints must tolerate and what substrate the rest of the cabinet should be built from.

Continue Learning

Planning a kitchen and want to talk it through with the person who will build it?

Request a Design Consultation