When a series of patients' urine samples supplemented in vitro with chlorpromazine or imipramine was assayed with the Forrest qualitative assays, we observed an occasional false-negative result, which we found was attributable to interference by ascorbic acid. It interferes with the reagent, not with the analytes, in both assays. We easily eliminated this interference with the phenothiazine test by using an anion-exchange resin. Eliminating the interference with the assay for imipramine, however, is more difficult; false-negative results can be obtained even after ion-exchange chromatography if the imipramine concentration is less than 50 mg/L.