Stopped Dado Joint — What It Is, Where the Strength Comes From, and the Numbers

A stopped (or blind) dado is a standard dado joint that ends short of the front edge instead of running all the way across the board — the same bearing shoulder and registration a through dado offers, without the horizontal line showing on the case's front edge.

The one problem a router leaves that a saw doesn't

A dado cut with a table saw's dado stack necessarily runs the board's full width — there's no way to stop a saw blade mid-cut and lift it out cleanly. A router, by contrast, can plunge in from the front edge and stop wherever the layout calls for it — but it leaves a rounded end matching the bit's own radius, because a round cutter can't produce a square inside corner. That round end has to be dealt with one of two ways: squared off by hand with a chisel, or left round and matched by notching the front corner of the shelf itself to clear it.

Which fix to choose, and why it isn't arbitrary

Squaring the stopped end by hand keeps the shelf's front edge a plain rectangle — simpler shelf-making, more chisel work on the case side. Notching the shelf's corner to match the router bit's radius keeps the case side's channel untouched after routing — faster on the case side, but now every shelf needs an individually matched notch cut into its front corner, which adds up fast on a case with many identical shelves. Most shops pick one default and stick with it across a project rather than mixing approaches, since a mixed approach means remembering which method applies to which shelf.

How far short of the front edge to stop

The dado is typically stopped somewhere between 1/2" and 1" back from the front edge — enough that the housed shoulder is fully hidden behind the case's face frame or front lip once assembled, without stopping so far back that the shelf's own front corner is left unsupported by the channel for a visually awkward distance.

What a stopped end costs in bearing area, and why it rarely matters

Stopping a dado short of the front edge technically removes a small amount of the channel's total shoulder length compared to a through dado of the same depth — on a typical case side, the lost length is only the inch or so set back from the front, a small fraction of a shelf's full depth. Because a shelf's load is spread fairly evenly across the whole bearing shoulder rather than concentrated at the front edge, that small reduction almost never changes the joint's practical carrying capacity; the stopped dado's real cost is entirely in the extra layout and finishing work at the stopped end, not in any meaningful loss of strength.

Cut sequence

1. Mark the channel's full layout, including exactly how far short of the front edge it should stop.
2. Rout the channel from the back of the case toward the front, plunging and stopping precisely at the marked line — a router's edge guide and a clamped stop block both help hold that stopping point consistent across several identical dados.
3. Square the stopped end with a chisel, or leave it round and prepare to notch the mating shelf instead.
4. If squaring by hand, pare carefully to the router-cut walls' existing line rather than trying to establish a fresh line freehand.
5. Cut the shelf to fit, notching its front corner if the round-end approach was chosen.
6. Dry-fit and confirm the shelf's front edge lands flush with the case's front, not proud or recessed.

Why this is worth the extra step over a through dado

On any case piece with a visible front edge — most cabinetry, most bookcases, most furniture where the front of the piece is meant to look clean — a through dado's horizontal witness line at every shelf is a design compromise most makers would rather avoid. The stopped dado's extra hand work at each channel's end is a small, one-time cost per shelf location in exchange for that cleaner front edge on every piece the case ever shows.

Weighing the router's extra plunge-and-stop time against a through dado's single pass

The router's plunge-and-stop technique and the hand-squaring (or shelf-notching) that follows both add real per-channel time next to a single full-width dado-stack pass. On a case with several shelves cut at the same height, the setup for stopping consistently pays for itself quickly across every repeat; on a one-off shelf, it's a modest but genuine extra step with no repeat cuts to spread the setup cost across.