Long, usually tapered, and almost never the rectangle people measure it as. The width changes along its length, and that is where most of the error comes from.
Driveways flare at the road, widen at the parking area, and narrow past the corner of the house. Measuring once in the middle and multiplying by the length can be out by a large margin.
Take the width at intervals along the run, every ten to fifteen feet on a long drive, and average them. Better still, treat obviously different sections as separate rectangles: the approach, the straight run, the parking or turning area at the top.
The flare at the road end is worth measuring properly rather than estimating. It is often the widest part of the whole drive and it is triangular rather than rectangular, so treating it as full width for its whole length over-orders significantly.
A long straight drive is exactly what a laser measure is good at, provided there is something at the far end to shoot. A garage door or the house wall works perfectly.
Pointing down the drive toward the road is the awkward direction, because there is usually nothing there. Prop a board or a target plate at the end, or measure from the road toward the house instead and use the building as your target. Measuring in the direction where the solid object is saves a lot of frustration.
Sun makes it harder again, and a driveway is an outdoor daytime measurement almost by definition. The outdoor picks are here, and the general outdoor limitations are covered here.
This is the part that separates a driveway that lasts from one that ruts within two seasons.
A driveway is a structure: a compacted sub-base carrying the load, sometimes a binding layer, then the surface you see. Each layer has its own depth and each is a separate volume calculation over the same area.
Calculate them individually and add. Area times depth in inches, divided by 324, gives cubic yards for each layer. The conversion is here.
Actual depths depend on the ground beneath, the drainage and what will drive on it. A drive taking a delivery lorry needs more than one taking a single car. This is a specification question for your supplier or contractor rather than a number to take from a generic article, and it is worth asking rather than guessing, because under-building the base is the most expensive mistake available here.
Aggregate layers are laid loose and compacted down, so the volume you order is greater than the finished depth suggests. Suppliers can give you a compaction factor for the specific material. Skipping this is how people end up half a load short with the machinery already booked.
Measure the height difference between the top and bottom of the drive. This matters for two reasons.
Drainage is the obvious one. Water runs downhill, and if the drive falls toward the house or the garage you have a problem to design around rather than a detail to ignore. Where it falls toward the road, that is usually fine, though you may still need a channel drain at the boundary depending on local requirements.
The less obvious one is that a sloped surface is larger than its plan area, exactly like a roof. On a gentle drive the difference is negligible. On a genuinely steep one it is not, and the same pitch multiplier logic applies. The multiplier table works for any slope.
Measuring the fall over a long distance is a natural indirect measurement if you have a tool with an angle sensor. Covered here.
Loose surfaces need edge restraint, or the material migrates into the lawn and the drive gets progressively narrower and thinner. That is a linear measurement around the perimeter, not an area one.
Measure both sides of the run plus any exposed end. Where the drive abuts a wall or a kerb, that side needs nothing. Where it meets grass or a bed, it does. Gravel depths and quantities are covered here, and patios follow similar layered logic.
If none of the above is quite right, the eight we recommend most often are on our front page.