A number of patients have experienced treatment failure while receiving non-nucleoside reverse transcriptase inhibitor (NNRTI)-based antiretroviral therapy (ART), particularly in resource-limited countries. The need remains for clinical data on protease inhibitor (PI)-based regimens in these patients. A retrospective cohort study was conducted among HIV-1-infected patients who had failed NNRTI-based regimens, were naive to protease inhibitors (PIs), and subsequently initiated a salvage PI-based regimen between January 2004 and December 2006. The study period ended on 30 December 2007. One hundred and forty patients received a single-boosted PI +/- optimized background regimen (OBR) and 64 received double-boosted PIs. The median (IQR) duration of follow-up was 19 (13-29) months. The overall virological failure rate at 24 months was 15.2%. No statistically significant difference was detected between the two regimen groups (single-boosted PI +/- OBR 16.4% vs. double-boosted PIs 12.5%, log rank p = 0.818). At the end of the study, the median (IQR) change in CD4 cell counts for patients in the double-boosted PI group was higher than for patients in the single-boosted PI +/- OBR group [149 (53-322) vs. 105 (23-199), respectively, p = 0.012]. Patients receiving double-boosted PI regimens displayed a higher frequency of hypertriglyceridemia than those patients who received a single boosted PI +/- OBR (31% vs. 11%, respectively, p = 0.001). Boosted PI-based regimens showed acceptable virological outcomes among patients who had failed NNRTI-based ART. In the subgroup analysis, patients who received double-boosted PIs demonstrated a superior immunological response but not better virological outcomes compared to the single-boosted PI +/- OBR group.