Initial viral decay rate may be useful when comparing the relative potency of antiretroviral regimens. Two hundred twenty-seven ART-naïve patients were randomized to receive efavirenz (EFV) (n = 74), lopinavir/ritonavir (LPV/r) (n = 77), or atazanavir/ritonavir (ATV/r) (n = 79) in combination with two NRTIs. The most frequently used NRTI combinations in the EFV and ATV/r groups were the nonthymidine analogues tenofovir and emtricitabine or lamivudine (70% and 68%, respectively) and, in the LPV/r group, lamivudine and the thymidine analogue zidovudine (89%). HIV-1 RNA was monitored during the first 28 days after treatment initiation. Phase 1 and 2 decay rate was estimated in a subset of 157 patients by RNA decrease from days 0 to 7, and days 14 to 28. One-way ANOVA and subsequent Tukey's post hoc tests were used for groupwise comparisons. Mean (95% CI) HIV-1 RNA reductions from days 0 to 28 were 2.59 (2.45-2.73), 2.42 (2.27-2.57), and 2.13 (2.01-2.25) log(10) copies/ml for the EFV-, LPV/r-, and ATV/r-based treatment groups, respectively, with a significantly larger decrease in the EFV-based group at all time points compared with ATV/r (p < 0.0001), and with LPV/r at days 7-21 (p < 0.0001-0.03). LPV/r gave a greater RNA decrease compared with ATV/r from day 14 (p = 0.02). Phase 1 decay rate was significantly higher in the EFV group compared with LPV/r (p = 0.003) or ATV/r (p < 0.0001). No difference was found in phase 2 decrease. EFV-based treatment gave a more rapid decline in HIV-1 RNA than did either of the boosted protease inhibitor-based regimens. The observed differences may reflect different inherent regimen potencies.