Last Updated on October 20, 2026 by Umar Farooq
How to test soil compaction before a paver base comes down to three moves: walk it, probe it and then run a compactor over it. Well-compacted subgrade barely takes a heel print, a steel rod pushes in the same shallow distance everywhere, and the compactor stops sinking and starts bouncing. Anything that ruts, squishes or lets the rod drop an inch deeper than its neighbors is a soft spot, and it will be a dip in the patio by next spring. Labs and nuclear gauges give you a percentage, but for a backyard patio these three checks, done in the right order, tell you almost everything.
This guide is about checking the result. If you still need to know how to get soil dense in the first place, start with our guide to compacting soil, then come back here to prove it worked.
How to test soil compaction: what "compacted enough" means
Engineers do not say "feels firm". They say a percentage of the maximum density that particular soil can reach. For pedestrian areas and residential driveways, the CMHA guide specification asks for the subgrade to reach at least 98% of standard Proctor density (ASTM D698) (see CMHA PAV-GSP-011-21). The Proctor test is a lab procedure that finds the densest a given soil can get at its best moisture, and 98% means you got within two points of that ceiling.
You will not run a Proctor test on a 12 x 14 foot patio. What you can do is test for the same thing it measures: resistance to movement under load. The field checks below are cruder than a lab, but they all answer the same question, which is whether this ground will keep shrinking under the base after you leave.
Start with a heel print and a walk-over
Do this first because it costs nothing and tells you where to look harder.
- Let the dig sit until the soil is damp but not muddy. Soil that is too wet will feel springy even when it is fine, and soil that is bone dry will feel hard when it is not actually dense.
- Walk the whole excavation in your normal shoes. Note any spot that gives, squelches or leaves a deep print.
- Stomp your heel once on a doubtful spot. Firm ground leaves a faint mark. Weak ground leaves a crater with a rim.
- Look for water. A patch that stays darker, shinier or wetter than its surroundings is draining badly or sitting on something soft.
Mark every doubtful spot with a flag or a bit of spray paint. You are not fixing anything yet, just building a list.
Probe it with a rod and compare the depths
A steel rod, a long screwdriver or a section of rebar turns "feels firm" into something you can compare. The method is simple: push it straight down with steady hand pressure and note how far it goes in. The exact number matters less than the pattern, because no rod test gives you a percentage. What it gives you is a map of where the ground is softer than the rest.
Lay out a grid roughly every 3 to 4 feet, add points along the edges, and add any spot you flagged on the walk-over. Use the same rod and the same push every time. If eight points stop within an inch or two of each other and the ninth drops several inches deeper, that ninth point is your problem.
Two things will fool you. A buried stone stops the rod early and looks like a super-dense spot, so probe again an arm's length away before you trust a hard reading. And a rod pushed into dry topsoil will not tell you anything useful about the clay under it, so test in the layer you are actually building on, not the loose surface.
Let the compactor grade its own homework
The most honest test of whether soil is compacted is whether compacting it any further does anything. Run your plate compactor over the area and watch what happens.
- Still sinking and leaving a dish behind: the soil is loose or too wet. Keep going, or stop and let it dry.
- Machine travels easily and the surface barely changes between passes: you are done.
- Surface ripples or squeezes up ahead of the plate: the soil is too wet, or there is a soft layer underneath. More passes will not fix that.
- Plate bounces and the engine note rises: the ground is hard enough that the plate is rebounding, which is what you want to hear.
A forward-only plate is fine for the final check on a patio-sized area. The compactor size matters more for the base lifts that come next, which depend on the plate size. If you do not own a plate, renting one for a day is the cheapest test you can buy, and our guide to plate compactor rental covers sizes and prices.
Proof rolling is the test the road crews use
Proof rolling means driving something heavy across the finished subgrade and watching what it does. The Geo-Institute describes it as a way to locate soft, yielding or pumping material so it can be fixed before more fill or pavement goes in, and notes that it covers large areas but is observer-dependent and "not a precise test" (see its proof rolling summary).
Highway agencies get specific. One DOT specification summary calls for a tandem-axle dump truck loaded to at least 20 tons, one pass per lane, a maximum deflection of 1 inch on new subgrade, and no pumping or cracking (see the proof rolling abstract on ROSAP). You are not hiring a 20-ton truck for a patio, and the homeowner version is smaller: a loaded wheelbarrow, or the plate compactor itself, doing the same job of finding the spot that gives.
That is the useful idea to borrow. Do not test the soil with a feather (a patio that passes a feather test fails the first barbecue). Test it with something at least as heavy as what it will carry, and watch for the place that moves.
When a lab or gauge test earns its fee
For the great majority of patios and walkways, paying for a density test is overkill, and I would say so even to a reader who offers to pay. The one-day cost of a rented plate and an hour with a rod is the better spend. But a few situations justify a number:
- A paver driveway or RV pad, where the load is heavy and the base is deep (see our guide to building a paver driveway).
- Fill soil you brought in or built up more than a few inches, where the ground was not undisturbed to begin with.
- A permit or an HOA that asks for a test report.
- A site with a known history of settling, such as a backfilled trench or a filled-in pool.
The usual field methods are the sand cone, the nuclear density gauge and the dynamic cone penetrometer. The first two measure density directly. A geotechnical firm or a local testing lab runs them, compares the result to the Proctor figure for your soil, and gives you a pass or fail. Ask for the target to be 98% of standard Proctor for a pedestrian or residential driveway surface, matching the CMHA specification above.
What to do with the soft spots you found
A soft spot does not mean the whole dig is ruined. Fix it in this order:
- Check moisture. If the spot is simply wet, wait a day or two, tarp it against rain and test again.
- Dig it out. If it is still soft once dry, remove the weak soil down until the rod meets firm ground.
- Refill in thin layers. Put back crushed stone or soil in lifts a few inches thick and compact each one. Our guide to how to compact gravel covers the lift-by-lift method.
- Retest. Probe the repair against its neighbors. It should now read the same as the rest.
If the ground is soft over most of the site, the answer may be a deeper base or reinforcement rather than more passes. ICPI Tech Spec 2 adds 2 to 4 inches to the base on continually wet or weak soils, and our paver patio on clay soil article covers what to do when the ground is weak by nature.
Firm, even and the same everywhere is a pass
The test is not a single number but a comparison. A good subgrade barely marks under a heel, takes a probe to roughly the same depth in every spot, and stops changing under the plate compactor. If all three agree, put the base on it. If one disagrees, believe the one that found the problem. Dirt is honest, it just takes a spring to say so.
FAQ
How can you tell when soil is compacted enough?
It resists your heel, it takes a probe rod to about the same depth everywhere, and a plate compactor stops leaving a dish behind. For a formal answer, the CMHA guide specification asks for at least 98% of standard Proctor density under pedestrian areas and residential driveways, which only a lab or gauge test can confirm.
What is the easiest way to test soil compaction at home?
Probe it. Push a steel rod or long screwdriver into the soil at the same pressure at points about every 3 to 4 feet and compare how far it goes. Reading that is noticeably deeper than the rest marks a soft spot. It will not give you a percentage, but it finds the weak places.
What is a proof roll?
It is a test where a heavy loaded vehicle or roller crosses the finished subgrade while someone watches for ruts, deflection or pumping. Highway crews use loaded dump trucks. For a patio the equivalent is running a plate compactor over the area and watching for ground that moves.
Should I test soil when it is wet or dry?
Test it when it is damp but not muddy. Wet soil feels springy and fails the heel test even when it is fine, and very dry soil feels hard without being dense. If it is wet, tarp it and wait for it to dry before you judge it.
Do I need a density test for a paver patio?
Usually not. Checking the ground with a probe, a walk-over and a compactor is enough for most patios and walkways. A density test is worth paying for under driveways and RV pads, on deep fill, or where a permit calls for it.
What if my soil fails the compaction test?
Fix the soft spots first: let wet ground dry, dig out anything that stays soft, refill in thin compacted layers and retest. If the whole site is weak, plan a deeper base. ICPI adds 2 to 4 inches on continually wet or weak soils.
