Urine toxicology is the gold standard for estimating the prevalence of illicit drug use in methadone maintenance treatment (MMT). The frequency of urine testing may be crucial for establishing accurate use rates. Infrequent testing may lead programs to undercount active drug users and to target interventions too narrowly. This study compared results from frequent testing (twice per week) versus less frequent testing of 166 patients at four MMT programs. As part of a research study, all patients were tested by research staff for opioid and cocaine use twice per week on a fixed schedule for 10 weeks. During the same period, the four MMT programs tested the patients according to their standard protocols, approximately weekly (one program) or every 3-4 weeks (three programs). The research tests identified approximately 50% more illicit opioid users and 70% more cocaine users than the less frequent program tests. Patients who were drug positive according to the research tests but drug negative according to the program tests tended to be infrequent users. The data suggest that standard urine testing practices in MMT programs may result in underestimates of the prevalence of opioid and cocaine use. More frequent testing, even for time-limited periods, should produce more accurate depictions of drug use prevalence and help indicate the direction of interventions.