DOT Random Drug and Alcohol Testing Rates for 2026 (FMCSA)

For 2026, FMCSA requires you to randomly drug test at least 50% of your average number of CDL driver positions and randomly alcohol test at least 10%. These are the same rates FMCSA has used since 2020. The DOT Office of Drug and Alcohol Policy and Compliance (ODAPC) lists them in its annual random testing rate chart.

The 2026 FMCSA random testing rates

FMCSA can raise or lower these rates each year based on industry-wide violation rates. Any change is published in the Federal Register and takes effect January 1 of the next year. When the rates don't change, ODAPC notes this on its random testing rates page. For 2026, ODAPC shows the FMCSA rates as unchanged.

These are minimums. You may test at a higher rate if you choose. You may not test at a lower one.

Who is in the random pool

Part 382 testing covers drivers who operate commercial motor vehicles that require a CDL. If a driver needs a CDL to drive for you and performs safety-sensitive functions, that driver belongs in your FMCSA random pool. See 49 CFR 382.103 for who is covered and the few exceptions.

Keep the pool current:

• Add new drivers once they are hired and have passed their pre-employment test.
• Remove drivers who leave.
• Include part-time, seasonal and casual CDL drivers who perform safety-sensitive work.
• Don't put non-covered workers in the FMCSA pool. Under 382.305(j)(2), if you use a service agent's larger pool, only covered employees may be in it.

How the rate is applied to your driver positions

The percentage is applied to your average number of driver positions, not to the number of drivers on your roster on one particular day. This matters if your fleet grows or shrinks during the year.

Under 382.305(j)(1), you figure the average like this:

1. Count the covered drivers eligible for random testing in each selection period.
2. Add those counts together.
3. Divide by the number of selection periods in the year.

If you select more often than monthly, you do not have to compute this more than once a month.

Example

Say you select quarterly and your driver count was 8, 10, 12 and 10 in the four quarters. The total is 40. Divide by 4 periods and your average is 10 driver positions.

The number of tests is what counts, not the number of different people tested. Some drivers may be tested more than once in a year, and some may not be tested at all. That is normal with a truly random method.

Scientifically valid selection

49 CFR 382.305(i) sets three rules for how you pick drivers:

• Use a scientifically valid method. Examples in the rule are a random number table or a computer-based random number generator matched to driver identifiers, such as employee or payroll numbers.
• Give every driver an equal chance each time. A driver tested last quarter goes back into the pool and has the same chance of being picked again.
• Test every driver who is selected during that selection period. If a selected driver is on leave or off for the whole period, document why the test did not happen.

What is not allowed is picking names by hand, skipping a driver because they are busy, or choosing drivers you suspect of something. Testing based on suspicion is a separate test type with its own rules.

Most small carriers have a consortium or C/TPA run selections with software. That works, but the carrier is still responsible. Under 382.305(j)(2), you must make sure your service agent tests at the right rate.

Spreading tests across the year

Under 382.305(k):

• Random tests must be unannounced.
• Test dates must be reasonably spread throughout the calendar year.

In practice, this means you should not do all your random tests in December to catch up. Many carriers select quarterly or monthly so tests happen all year. Varying the day of the week and time of day also helps keep tests truly unannounced.

Alcohol tests have an extra timing rule. Under 382.305(m), a random alcohol test can only be done while the driver is performing safety-sensitive functions, just before, or just after.

The "immediately proceed" rule

Under 382.305(l), once you notify a driver that they were selected, the driver must go to the test site immediately.

There is one narrow exception. If the driver is performing a safety-sensitive function other than driving when notified, such as fueling or loading, they must stop and go to the test site as soon as possible.

Good practice for the notice:

• Notify the driver in person or by phone, not by a message they might see later.
• Record the time you notified them and the time they arrived at the collection site.
• Do not give advance warning, such as "you'll be tested next week."

A driver who is told to test and does not show up within a reasonable time may be treated as refusing. A refusal has the same consequences as a positive test. See our guide on what counts as a refusal to test.

Owner-operators and consortium pools

If you are an owner-operator with your own authority, you are both the employer and the driver. You cannot run a fair random program with a pool of one.

49 CFR 382.103(b) says an employer who employs only himself or herself as a driver must have a random testing program with two or more covered employees in the selection pool. 382.305(j)(2) lets you use a service agent and put your driver in a larger pool.

For most owner-operators, the answer is to join a consortium. A consortium, or C/TPA, combines drivers from many small carriers into one pool, runs the random selections, and tells you when you are picked.

When choosing a consortium, confirm that:

• The pool includes only DOT-covered drivers at the FMCSA rates or higher.
• You get written proof of enrollment.
• They provide selection records and your annual summary.
• They can act as your designated C/TPA in the Clearinghouse if you need it.

Owner-operators also have Clearinghouse duties. See our Clearinghouse queries guide.

Record-keeping for random testing

Under 49 CFR 382.401, you must keep testing records for set periods. The main ones for random testing are:

For random testing, an auditor will usually want to see:

• Your pool roster for each selection period
• The list of drivers selected and the date of each selection
• Proof each selected driver was tested, or a documented reason why not
• Your annual summary showing you met the 50% and 10% rates

If you use a consortium, ask for these documents each year and keep your own copies. "My consortium has it" is not a strong answer during an audit.

Why this matters in a new entrant audit

Failing to set up a random testing program is one of the automatic failure items in the table in 49 CFR 385.321. One occurrence is enough to fail. Our guide on new entrant audit automatic failure violations covers the full list, and our new entrant safety audit guide explains the audit itself. To find out when your audit is likely to happen, try our new entrant audit date tool.

Quick checklist for 2026

• Confirm your consortium or C/TPA is using 50% for drugs and 10% for alcohol.
• Make sure every covered CDL driver is in the pool, and only covered drivers.
• Select at least quarterly using a scientifically valid method.
• Notify selected drivers without warning and send them immediately.
• Track your average driver positions and tests completed.
• Get selection records and an annual summary, and keep them for the required time.

How JMM Compliance can help

We can enroll your drivers in a random testing pool, check that your selections meet the 2026 rates, and set up the records an auditor will ask for. If you are an owner-operator, we can help you join a consortium and complete your Clearinghouse setup. We are not a lab or an MRO, and we work with qualified service agents for collections and test results. Learn more on our random drug testing service page or our drug and alcohol program service.

This page is general information, not legal advice. Requirements depend on your operation; confirm details with FMCSA or a qualified professional.