Addressing the consistency among adolescents’ self-reported usage of marijuana, qualitative on-site urine screening, and quantitative laboratory testing, this article evaluates the drug-testing component of the Cannabis Youth Treatment (CYT) study. Noting that the prevalence of cannabis use by eighth graders in the United States more than doubled from 1991 to 1999, the authors maintain that these numbers may be low because of many adolescents denying substance use, underreporting the amounts of drugs used, and exaggerating the time intervals between drug usage. Focusing on data collected from 248 CYT program adolescents, aged 12- to 18-years-old, from 5 substance abuse out-patient treatment programs at 4 sites in Florida, Pennsylvania, Connecticut, and Illinois, this cross-sectional study evaluated the consistency and validity of self-reported drug use, on-site urine testing, and laboratory urine testing. Analyzing kappa coefficients, the authors found that adolescents’ self reported drug use rates were higher at intake than either of the two urine tests but were both lower and higher at the 3-month and 6-month follow-up studies. The authors determined that the statistical disagreements went in both directions, with the kappa coefficients in the moderate range only. Furthermore, the authors found that over two-thirds of the frequent cannabis users tested positive for marijuana when they claimed they had not used drugs for 1 week and over one-third tested positive after claiming that it had been more than 4 weeks since they used cannabis. The authors conclude that future studies need to assess which people are most likely to have inconsistent answers when comparing self-reported drug use with laboratory and on-site urine testing. Tables, references