Random Drug Testing Programs

Why Random Programs Fail

Random testing is the most commonly run and most commonly botched part of an employer testing program. Not because employers act in bad faith, but because the work is administrative and never urgent, so it slips.

The pattern is familiar. The pool was built when the program started and has not been reconciled against payroll since. Selections happen when someone remembers. A supervisor quietly skips a name because the person is mid-project. Nobody can produce documentation of how the last four draws were generated.

Each of those is survivable on its own. Together they mean that when an employee tests positive and challenges it, the company cannot demonstrate the selection was random, and the whole program looks arbitrary. That is the risk a well-run random program is designed to remove.

What Gets Handled

• Pool management. The covered population is maintained against your roster: new hires added, departures removed, role changes moved between pools.
• Selection. Draws are generated on your schedule by a method nobody at your company can influence, and each draw is documented.
• Notification. Your designated contacts get the selections, with the process and timing your policy specifies for getting the employee tested.
• Collection. At a collection site, or for non-regulated testing, remotely with live observation, which is what makes random testing practical for remote and multi-site workforces.
• Results. Negatives reported; non-negatives confirmed by laboratory and reviewed by a Medical Review Officer before any verified result reaches you.
• Records. Selection documentation, completion tracking, and rate performance, so you can show the program ran as written.

Setting the Rate

Most employers express the rate as a percentage of the pool tested per year, spread across draws through the year rather than done all at once. Quarterly draws are common because they keep selection unpredictable while remaining administratively sane.

Two rules matter more than the number you pick:

1. Write it in the policy and meet it. A rate you routinely miss is evidence the program is not what you say it is.
2. Apply it to the whole pool. Quietly excluding a department, a shift, or a location is the fastest way to turn a testing program into a discrimination claim.

Multi-Site and Remote Workforces

Random testing gets harder as people spread out. A selection at a single plant is a walk down the hall. The same selection across eleven sites in six states means eleven local arrangements, inconsistent turnaround, and a real chance that a selected employee simply never gets tested.

Two things fix most of it: one program that makes selections centrally rather than site by site, and collection options that reach people where they are, including remote collections for non-regulated testing when an employee is nowhere near a collection site.

Recordkeeping That Holds Up

If your program is ever questioned, in an arbitration, an unemployment hearing, or a lawsuit, the questions are predictable: who was in the pool, how were names chosen, who ran the selection, was anyone excluded, and was the person tested promptly after notification. A program that can answer those in writing is in a completely different position from one relying on memory.

Random testing is rarely bought on its own. It usually arrives as part of wider employee drug testing services, alongside pre-employment and post-accident coverage.

Where Random Testing Works Differently

Random is the part of a national program most likely to be unlawful in a particular state, because several states regulate the draw itself rather than the decision to test:

• Vermont prohibits random and company-wide testing outright except where federal law requires it, with no safety-sensitive carve-out and no way to cure it by agreement or consent.
• Connecticut requires the Labor Commissioner's written approval before an employer may begin random urinalysis. A job title appearing on the state's published safety-sensitive list is not authorization.
• Iowa dictates the mechanics: selection by an entity independent from the employer, using a computer generator matched to employee identifying numbers, with every attempt logged. Getting it wrong carries $1,000 per violation.
• Montana requires all supervisory and managerial employees to be in the pool, and requires the testing rate to be fixed in advance and stated in the policy.
• Minnesota limits random testing to genuinely safety-sensitive positions and to CBA-covered professional athletes.
• South Carolina conditions its premium credit on random sampling of everyone receiving wages in any form, which a risk-tiered pool does not satisfy.

The pattern worth noticing: a pool design that is standard nationally can fail in one state on a detail nobody at the company chose. Our state law pages set out each one with the statute it comes from.