In a 6-month randomized trial comparing 125 opiate-dependent patients who were assigned to four treatment groups (2 or 6 mg of buprenorphine and 35 or 65 mg of methadone), we examined the effects of cocaine use on opiate withdrawal symptoms measured on a 25-item scale on which the scores range from 0 to 75. For the methadone-maintained patients receiving the relatively low dose (35 mg), weekly withdrawal symptoms were highest when the urine toxicology for that week indicated no cocaine use. Similar associations were found for buprenorphine. Thus, when using cocaine at a low maintenance opiate dose, persistent opiate withdrawal symptoms were reduced, which is consistent with previous naloxone-precipitated withdrawal studies. Interestingly, with a higher dose of buprenorphine (6 mg), cocaine may have increased opiate withdrawal symptoms, suggesting a possible mechanism for the reduction of illicit cocaine abuse also recently observed in another study in patients treated with high dose (120 mg) methadone maintenance. This has led to a two-component model for the relationship between cocaine and opiate withdrawal-like symptoms at high versus low opiate maintenance dose. This two-component model also reconciles the contradictory findings of prior studies.