How to Set Up a Random Drug Testing Program

Step 1: Define the Pool

Decide who is covered: everyone, or every employee in a safety-sensitive role. Define safety-sensitive by task rather than title, so the definition survives reorganizations and title changes.

Then decide how the pool stays current. This is the part that quietly fails. New hires have to be added, departures removed, and role changes moved between groups. A pool reconciled against payroll once a year is wrong for most of that year, and selections drawn from it include people who left.

Step 2: Set a Rate You Will Meet

Express it as a percentage of the pool tested per year. Regulated programs have rates set for them; everyone else chooses. Base yours on risk and on what you can sustain, then write it in the policy.

Two rules matter more than the number. Apply it to the whole pool, with no quiet exclusions for a department or a shift. And actually meet it, because your own completion records are the first evidence anyone examines.

Step 3: Make Selection Genuinely Random

Every person in the pool must have an equal chance in every draw, generated by a method nobody inside the company can influence. That rules out picking names, rotating through a list, and skipping someone because it is a bad week for their department.

Document each draw: when it ran, how many were selected, who was in the pool at the time, and what method produced the result. That documentation is the whole defense if the program is ever questioned.

Step 4: Set the Notification Window

Say how soon after notification an employee must report for testing. A long gap defeats the purpose, and an undefined gap produces inconsistency between managers. Also decide who notifies, what they say, and what happens if the employee refuses or delays.

Step 5: Keep the Records

At minimum: pool membership over time, selection documentation for each draw, completion tracking including refusals and non-completions, and rate performance against your policy. If asked next year who was in the pool last quarter and how names were chosen, you want that to be a retrieval rather than a reconstruction.

Announcing It Without Alarming People

Give notice well before the first draw. Explain why the program exists, who is covered, how selection works, and that repeats can happen. Train supervisors before employees hear about it, because they will be asked first and a confident answer prevents most of the anxiety.

The Failure Pattern to Avoid

It is always the same: the program starts strong, the pool is not maintained, a quarter gets skipped when someone is busy, a site is quietly left out because testing there is inconvenient, and two years later the company cannot demonstrate that selection was random. Nothing dramatic happens until the day it matters, and by then the record is what it is.