Drug users can be highly motivated to obtain negative results on urine drug tests and may attempt to subvert the process by in vivo adulteration. The use of herbal products for "flushing" and "detoxification" is frequently advertised as an effective means of passing drug tests. Accordingly, a study was designed to determine the effects of ingestion of two herbal products, Naturally Klean Herbal Tea and Golden Seal root, and a diuretic medication, hydrochlorothiazide. The herbal tea was prepared in 1 gal of water as specified by the manufacturer. All other products were consumed with 1 gal of water. Two control conditions in which the subject consumed only water (1 gal; 12 oz) were included. The 1-gal liquid treatments were divided into 4-qt aliquots, and 1-qt was consumed each hour for 4 h. All treatments were begun approximately 22 h after smoking of a marijuana cigarette (3.58% THC) and 22 h after intranasal administration of cocaine hydrochloride. Following all treatments with excess fluid, creatinine and specific gravity dropped in 1.5-2.0 h to levels indicative of diluted specimens (<20 mg/dL creatinine, <1.003 specific gravity). Marijuana and cocaine metabolite concentrations by immunoassay (EMIT and TDx) also dropped rapidly, and the results frequently switched from positive to negative. By the time subjects had consumed 2 qt of any fluid, they were generally producing false-negative results. For example, ingestion of excess water produced dilute specimens (<20 mg/dL creatinine; <1.003 specific gravity) in an average time plus or minus the standard error of the mean of 1.47 +/- 0.17 h (N = 5) and 1.45 +/- 0.2 h (N = 5) following smoked marijuana and intranasal cocaine, respectively. In comparison, ingestion of Klean Tea produced dilute specimens in 1.36 +/- 0.07 h (N = 4) and 1.39 +/- 0.11 h (N = 4) following marijuana and cocaine administration. Recovery of urine test measures to pre-treatment levels occurred over a period of 8-10 h. Average detection times for marijuana metabolite appeared to be slightly shorter following ingestion of 1 gal of fluids compared with ingestion of 12 oz of water as a result of the time of testing being near the end of the cannabinoid metabolite excretion phase. Consequently, negative cannabinoid results induced by fluid ingestion rarely returned to positive after excess water was eliminated. In contrast, negative cocaine results reverted to positive quickly after the dilution effects disappeared. It was concluded that excess water ingestion can produce false-negative test results, but the claims of herbal products to be an aid in passing a urine test appear to be unfounded.