Reports of prolonged drug excretion have provided the basis for the common assumption that cannabinoid metabolites may he detected in urine for a week or longer. The accuracy, sensitivity, and specificity of immunoassays for the detection of cannabinoids and metabolites are unique for a specific assay and may change overtime. it is important that individuals who select assays and those who interpret test results be aware of qualitative and quantitative changes that occur. In the present study, detection times of cannabinoids in urine were determined using cannabinoid immunoassays with 20-, 50-, and 100-ng/mL cutoffs and using gas chromatography-mass spectrometry (GC-MS). Six subjects each smoked a single marijuana cigarette (placebo, 1.75, or 3.55% delta9-tetrahydrocannabinol [THC]) each week while residing on the clinical ward of the Addiction Research Center. Each urine specimen was analyzed under blind conditions by immunoassay according to the manufacturer's instructions. The following cannabinoid reagents were evaluated: EMIT d.a.u. 100, EMIT d.a.u. 50, EMIT d.a.u. 20, EMIT II 100, EMIT II 50, Abuscreen OnLine, and Abuscreen RIA, DRI, and ADx. All urine specimens were also analyzed for 11-nor-9-carboxy-delta9-THC by GC-MS using a 15-ng/mL cutoff. Urinary cannabinoid detection times varied substantially across assays, subjects, doses, and cutoff concentrations. Detection times were shorter than previously assumed. Mean detection times increased from a maximum of 0.5 days after the low dose to 1.5 days after the high dose using the 100-ng/mL cutoff. Mean detection times were less than 1 day following the low dose and less than 2 days following high-dose exposure using the 50-ng/mL cutoff. Mean detection times ranged from 1 to 5 days after the low dose and from 3 to 6 days after the high dose using the 20-ng/mL cutoff immunoassay. GC-MS detection times were approximately twice as long as mean detection times using an immunoassay with a cutoff of 50 ng/mL. Differences in sensitivity and specificity between the available immunoassay products affected the efficiency of detection of marijuana use. These results indicate that recent reductions in cannabinoid cutoffs by military and federally mandated programs will increase detection times and improve sensitivity, as expected. However, monitoring acute marijuana usage with a commercial cannabinoid immunoassay that has a 50-ng/mL cutoff concentration provides only a narrow window of detection of 1-2 days.