OBJECTIVE To evaluate the associations between methadone and high-dose buprenorphine maintenance treatment and illicit drug use and injection among drug users in France. METHODS A cross-sectional study. Data were gathered using a questionnaire administered containing closed-ended questions. METHODS Drug dependence clinics (DDC) and general practitioners' (GPs) offices in three French cities. METHODS Drug users undergoing maintenance treatment with methadone (n = 197) and buprenorphine (n = 142). METHODS Interviews covered the use of illicit drugs (heroin, cocaine or crack) and injection practices (illicit drugs and/or substitution drugs) during the last month, current treatment modalities, socio-demographic and health characteristics. Bivariate analysis and multivariate logistic regressions were conducted. RESULTS Overall, 35.4% of respondents (34.5% in the methadone group, 36.6% in the buprenorphine group, P= 0.69) had used at least one illicit drug, 25.7% reported having injected drugs and 15.3% had injected the substitution drug. Injection was more common among buprenorphine-maintained individuals (40.1%) than among users on methadone (15.2%) (P < 0.01). Multivariate analyses indicate that the type of substitution drug (buprenorphine versus methadone) was not associated with illicit drug use (OR = 1.1; 95% CI = 0.7-1.8). In the buprenorphine group, injection was related independently to social situation, as measured by housing (unstable versus stable housing, OR = 4.3; 95% CI = 1.6-11.5), but this was not the case in the methadone group. The risk of injection increased with buprenorphine dosage (high/low dosage OR = 6.2; 95% CI = 2.0-19.7), but this association was not observed in the methadone group. CONCLUSIONS Further studies comparing the benefits of these two types of treatment should be carried out, taking outcomes such as physical health, mental health and social functioning into consideration.