The use of ritonavir as a protease inhibitor boost is rare in sub-Saharan Africa because a heat-stable formula is not available. We report the results of an open-label pilot trial with unboosted atazanavir in combination with lamivudine and didanosine as first-line therapy conducted in Senegal. Treatment-naive HIV-1 infected adult patients without active opportunistic disease were included. The primary endpoint was the proportion of patients with plasma HIV-1 RNA <400 copies/ml at week 48. Forty patients (12 men and 28 women; mean age +/- SD: 40 +/- 9 years) were included. Treatment was changed during the study for two patients (pregnancy, tuberculosis); one patient was lost to follow-up and one patient died (gastroenteritis with cachexia). At week 48, 78% [95% confidence interval (CI): 65-90%] and 68% (95% CI: 53-82%) of the patients had HIV-1 RNA <400 and <50 copies/ml, respectively (intent-to-treat analysis; not completer = failure). Among the seven patients with HIV-1 RNA >or=400 copies/ml at week 48, five were not compliant; genotyping analysis (n = 4) did not reveal a major mutation for protease inhibitors. The mean CD4 cell count change from baseline to week 48 was +238 +/- 79 cells/mm(3). The combination of unboosted atazanavir with lamivudine and didanosine was efficient and well tolerated in HIV-1-infected patients with results similar to those observed in Northern countries. These results suggest that unboosted atazanavir with its high genetic barrier could be a valuable alternative to NNRTIs in resource-limited countries in some HIV-1-infected patients in case of compliance issues with NNRTIs, intolerance to NNRTIs, resistance mutations to NNRTIs, in women with childbearing potential, or as a maintenance therapy in patients with virological suppression.