Numerous methods for the detection of urinary metabolites of marijuana constituents are available. Documentation of the sensitivity and specificity of these tests is needed before the determination of a pair of screening-confirmation tests can be made. This study used 88 clinical specimens to evaluate five commercially available marijuana metabolite methods and one new gas chromatography/mass spectrometry (GC/MS) method. The EMIT-d.a.u. test was found to have 2 to 3% unconfirmed positives when compared to the other methods evaluated. The new thin layer procedure, TOXI-LAB, was not as sensitive as the EMIT-d.a.u. procedure for some specimens, but proved to be a good confirmation for the EMIT-d.a.u. with elimination of all "unconfirmed positives." The Abuscreen (Roche) and the EMIT-st assays were positive for samples that contained larger amounts of 11-nor-delta 9-tetrahydrocannabinol-9-carboxylic acid (11-nor-delta 9-THC-9-COOH). The Immunalysis-radioimmunoassay (RIA) was positive for all samples found positive by the GC/MS method, but the concentrations found by the two assays did not correlate. The GC/MS method was developed to use the same extraction as the thin layer procedure and provides confirmation for all procedures except 2 to 3% of the positive EMIT-d.a.u. results.