SETTING: Nine countries in West and Central Africa. OBJECTIVE: To assess outcomes and adverse drug events of a standardised 9-month treatment regimen for multidrug-resistant tuberculosis (MDR-TB) among patients never previously treated with second-line drugs. DESIGN: Prospective observational study of MDR-TB patients treated with a standardised 9-month regimen including moxifloxacin, clofazimine, ethambutol (EMB) and pyrazinamide (PZA) throughout, supplemented by kanamycin, prothionamide and high-dose isoniazid during an intensive phase of a minimum of 4 to a maximum of 6 months. RESULTS: Among the 1006 MDR-TB patients included in the study, 200 (19.9%) were infected with the human immunodeficiency virus (HIV). Outcomes were as follows: 728 (72.4%) cured, 93 (9.2%) treatment completed (81.6% success), 59 (5.9%) failures, 78 (7.8%) deaths, 48 (4.8%) lost to follow-up. The proportion of deaths was much higher among HIV-infected patients (19.0% vs. 5.0%). Treatment success did not differ by HIV status among survivors. Fluoroquinolone resistance was the main cause of failure, while resistance to PZA, ethionamide or EMB did not influence bacteriological outcome. The most important adverse drug event was hearing impairment (11.4% severe deterioration after 4 months). CONCLUSIONS: The study results support the use of the short regimen recently recommended by the World Health Organization. Its high level of success even among HIV-positive patients promises substantial improvements in TB control.