OBJECTIVE: Because a centre effect can sometimes exist in HIV treatment, we sought to measure the heterogeneity of French hospital departments delivering HIV care and to test the presence of such an effect on adherence and response to highly active antiretroviral therapy. METHODS: The ANRS-EN12-VESPA study is a nationally representative 2-stage cross-sectional survey conducted in France in 2003 and covering 102 hospital departments providing HIV care. Each department described its HIV care activities and care provision. Analyses of adherence and 4 indicators of treatment outcome were restricted to the 699 patients diagnosed from 1996 onwards and treated with highly active antiretroviral therapy for at least 6 months. The variability between departments was assessed with random-effect models for binary outcomes. RESULTS: The departments delivering HIV care proved to be somewhat heterogeneous in numerous respects, including their size and their onsite provision of consultancies and other services, and the characteristics of their patient population. Mean observed adherence was 63.3%, and the means of the different treatment failure indicators ranged from 6.1% to 59.8%. The departments showed some variability for these outcomes, but no significant centre effect was detected. CONCLUSIONS: Despite the heterogeneity of the specific types of medical services offered by the hospitals providing HIV care, the nationwide treatment results seem homogeneous. This homogeneity could be attributed to the widespread and consistent application of therapeutic guidelines, which are regularly updated by consensus.