Detection windows vs impairment: why the gap costs jobs

Last verified: 2026-07-29

THC is detectable long after any effect has worn off. Workplace policies mostly test for detection, not impairment. Every strategy that ignores that gap eventually fails a test it did not expect to fail.

What tests actually measure

• Oral fluid (saliva) — governed by AS/NZS 4760:2019, screening for THC at a 15 ng/mL cut-off. Skews toward recent use, but “recent” can still be well after effects have faded, and regular use extends it.
• Urine — governed by AS/NZS 4308, detecting THC metabolites. Windows run to days, and to weeks in frequent users. It says something about history, almost nothing about now.
• Neither measures impairment. Unlike blood alcohol, there is no agreed THC number that equals impaired — a point tribunals themselves acknowledge.

Why “I wasn’t impaired” keeps losing

Because most policies do not claim to test impairment. They set cut-off limits and make exceeding them a breach. In Gauci v DP World, the Commission accepted the policy “was not based on levels of impairment” but on cut-offs that did not distinguish prescribed from recreational use — so the worker’s unimpaired state carried little weight. The Full Bench in Sharp v BCS reached similar territory in aviation: impairment could not be proven either way, and the reading against the cut-off was what mattered.

What this means practically

1. Know which test and which standard your workplace uses — it is usually in the policy or the testing procedure, and it changes your real-world margin enormously.
2. Timing and product choice are decisions, not luck. Dosing schedule relative to shifts, and THC content of the product, are exactly the things to work through with your prescriber — the Rights Kit has the question list.
3. Self-test kits are a guide, not a guarantee. Consumer kits may use different cut-offs than your workplace’s laboratory-confirmed process — the 2025 mining matter in our tracker turned partly on exactly that mismatch.