Introduction: We compared the effectiveness of tenofovir/emtricitabine (TDF/FTC) combined with either lopinavir/r (LPV/r) or another recommended third drug in the 2010 French guidelines in antiretroviral-naıve patients starting combination antiretroviral therapy in 2004Á2008 in the French Hospital Database on HIV. Methods: The outcomes were stop or switch of the third component, viral load (VL) B500 copies/ml, an increase of at least 100 CD4 cells/mm 3 , AIDS-defining event and non-AIDS-defining hospitalization or death. Propensity scores were estimated by logistic regression based on the clinical centre and other confounders. In each clinical centre, each patient initiating LPV/r was matched with a patient initiating another third drug (efavirenz or atazanavir/r) and having a close propensity score. Cox's proportional hazards models were then used, with treatment as covariate. Time was right-censored at four years. Results: 1269 patients started LPV/r plus TDF/FTC, and 890 could be matched to 890 patients receiving another third drug. Baseline characteristics were well balanced between these two groups. LPV/r was associated with a higher risk of third drug stop (hazard ratio (HR): 1.69; 95% confidence interval (CI), 1.42Á2.00) and with less rapid viral suppression (HR: 0.83; 95% CI, 0.72Á 0.95). There was no difference in the time required for a CD4 cell increment of at least 100/mm 3 , or to the occurrence of an AIDS-defining event. Non-AIDS-defining hospitalizations or deaths were more frequent with LPV/r (HR: 1.79; 95% CI, 1.33Á2.39). Conclusions: For first-line therapy, in this observational setting, TDF/FTC plus LPV/r were less durable than TDF/FTC plus another recommended third drug, led to a less rapid viral suppression and were associated with a higher risk of non-AIDS morbidity.