6 Cabinet Joinery Methods, and When Each One Actually Earns Its Place

Every cabinet box has to get joined somehow. Most shops use whatever method they learned first and never question it again, because it's worked fine for years. That's true right up until a box starts racking in six months, or a client asks for units that need to go together, come apart for a house move, and go back together again without falling apart. There's no single right answer here — each method below trades off speed, strength, cost, and whether the joint can ever come apart again. Pick wrong and you usually don't find out in the shop. You find out on the install, weeks later, when a drawer front sags or a corner opens up a hairline gap, and a callback to fix a joint costs a lot more than the extra two minutes it would've taken to build it right the first time.

Dado Joints Are Still the Shop Standard

A dado is a groove cut into one panel that the other panel sits into, usually about a quarter of the panel's thickness deep. It's been the default in small and medium shops for decades, run on a table saw, a router, or a CNC without much fuss. Cut once, and it locks the sides square before a single fastener goes in, which is a big part of why it's stuck around this long.

The groove does two jobs at once. It keeps the sides square while you glue and clamp, so the box can't twist out of true while the glue sets overnight. It also gives you a lot more glue surface than a plain butt joint — easily three or four times the contact area — and that wide glue line is really where the long-term strength comes from, not the groove itself.

The catch is precision. Cut a dado a hair too wide, or set it in the wrong spot, and the box won't close up square, full stop. There's no fixing that after the fact — you just re-cut the panel and eat the material and the time. On a CNC that's rare, since the program doesn't drift from part to part. On a hand-fed table saw with a worn fence, it happens more than anyone likes to admit, usually on the last panel of the day. It's also slower than most of the other options here, since every groove is a separate setup or pass, which is worth factoring into your quote before you commit to it on a tight-turnaround job.

Pocket Screws Are Fast, But Know Where They Belong

Pocket screws are quick. There's no clamping, no glue drying time, and no waiting around for anything to cure before you move the box on to the next step. A fitter with a jig can knock out a run of face frames in a fraction of the time a dado-and-glue approach would take, which is exactly why they're everywhere on site work.

On load-bearing carcass joints, fitters split into two camps, and neither one's wrong. Some won't screw a bottom straight into a side without a dado backing it up first, especially under a stack of dinner plates or a built-in appliance. Others run plain butt joints with pocket screws on kitchen boxes every single day and never see a failure, particularly when the screws land in solid ply rather than the raw edge of particleboard.

If you do use pocket screws on particleboard or melamine, watch your driver's torque setting closely. Drive too hard and you'll strip the hole, or blow the back right out of the board as the screw bottoms out. On melamine-faced stock that chip-out shows on the face, and there's no hiding it once the door's hung in front of a client.

Biscuits Help You Line Things Up, Not Carry Load

Biscuits are good at one specific job: lining panels up while you glue and clamp. The compressed wood wafer swells slightly once the glue hits it, which pulls the joint tight and stops parts sliding around mid clamp-up, especially on a long shelf edge where things like to creep out of line.

What biscuits aren't good at is carrying real weight on their own. Most shops that use them on a carcass still back the joint with glue and screws anyway, because a biscuit by itself in a shelf or bottom joint just isn't rated for that kind of load over time. Treat a biscuit as an alignment helper, not the actual joint, and you won't be disappointed by what it gives you. Where biscuits genuinely shine is on mitred corners and face-frame assemblies, where hiding end grain matters more than raw strength, and a biscuit joiner is a lot faster to set up than a dado stack.

Confirmat Screws Are the Easy Upgrade From Drywall Screws

A confirmat is a thick, coarse-threaded screw built specifically for particleboard and MDF edges, usually running 5mm or 7mm across and 40 to 50mm long. It bites into the material far harder than a standard wood screw or a drywall screw ever could in the same board, which is why it's become the default fastener in a lot of flat-pack and frameless work over the last couple of decades.

You can back a confirmat out and drive it back in a handful of times without losing much grip, which is rare for a fastener in particleboard. Most screws in that material are a one-shot deal once they're out. That makes confirmats a solid pick for cabinets that might get opened up and serviced years down the line, like a sink base that'll eventually need a new bottom panel after a slow leak finally shows itself.

They do need the right bit to work properly. A stepped confirmat bit drills the pilot hole, the counterbore, and the clearance hole in one pass, all sized to the exact screw you're using. Use a regular twist bit instead, and you'll fight the screw the entire way in, or crack the board trying to force it home, and then you're back to the supplier for another sheet.

Cam Locks Suit Knock-Down Work

Cam locks are the two-part fittings you'll recognize from flat-pack furniture: a barrel that sits in one panel and a cam that hooks into it from the other and tightens with a quarter turn. They let a cabinet go together, come apart, and go back together again without the joint wearing loose the way a plain screw hole eventually does after a few cycles of assembly and disassembly.

The trade-off is cost and time. Cam locks need two separate holes bored at the right depth and spacing on each panel, which takes longer than drilling for a dowel or driving a confirmat home. The hardware stays visible too, unless you cover it with a cap or hide it behind a panel. For units that need to ship flat, or get carried up a tight stairwell in pieces and assembled on site, that extra machining time pays for itself the first time the job would've been impossible any other way. They're also worth the hassle for anything a client might want to relocate themselves down the line, like a modular island, since the joint doesn't degrade the way a repeatedly-screwed hole does.

Dowels With Glue Are the Strongest, If You Have the Gear

Shop tests that circulate around the trade keep putting glued dowel joints ahead of pocket screws for raw strength, and well ahead of a plain butt joint with no reinforcement at all. The glue bonds along the full length of the dowel inside both panels, which spreads the load across a wide surface in a way a single screw point just can't match.

The catch is speed and setup. You need a doweling jig, or better, a dedicated boring machine, to drill clean holes that land in exactly the same spot on both panels every single time. Without that gear, dowelling a full kitchen's worth of boxes takes a lot longer than confirmats or pocket screws would, and a hole that's even a millimetre off throws the whole joint out of true.

For permanent installs where the box is never meant to come apart again, like a built-in that's fixed to the wall for good, dowels and glue are hard to beat. You're trading assembly speed for a joint that will genuinely outlast the rest of the kitchen around it, doors and worktop included. If you're running enough volume to justify the machine, that upfront cost pays back fast once dowelling is just another station on the line rather than a special job.

• What board you're joining — ply generally holds screws far better than particleboard does
• Whether the finished cabinet ships flat or gets carried up to site in pieces
• Whether it'll ever need to come apart later for a repair, a leak, or a swap
• What tooling you've actually got in the shop right now, not what you wish you had
• How much visible hardware the client will actually put up with on an open-back unit