We conducted a 6-week, single-blind study of acupuncture for cocaine dependence in methadone-maintained patients (N = 40) for the purpose of identifying an appropriate needle puncture control for use in future large-scale clinical trials. Patients were randomly assigned to receive daily acupuncture in three auricular sites plus one body site (LI-4), or in control sites within 2-3 mm of the four active sites. Overall, there was a positive response to treatment on a variety of drug-related and psychosocial measures. Cocaine use decreased significantly for patients in both needle puncture groups. The only statistically significant difference between the two types of needle puncture was on ratings of craving. Subjects rated each type of needle puncture as equally credible and perceived no significant differences on the acute effects of the two types of needle insertions. Power calculations based on these findings suggest that very large sample sizes would be required to detect treatment differences between this control condition and active acupuncture. Alternative controls are suggested, and the challenges inherent in implementing controlled clinical trials of acupuncture are discussed.