Objective: Efavirenz (EFV) was superior to nevirapine (NVP) in two recent cohort studies; but data from clinical trials suggest that three studies are needed to validate cohort results. We performed a cohort analysis comparing time to treatment failure and change in plasma HIV-1 RNA from baseline in antiretroviral therapy (ART)-naïve individuals treated with NVP- or EFV-containing regimens. Method: A cohort analysis of three observational databases (N = >10,000 patients) found 1,078 ART-naïve individuals treated with NVP-containing (n = 523) or EFV-containing (n = 555) regimens. Patients were evenly matched and received at least three antiretroviral agents. The primary endpoint was time to treatment failure defined as a rebound in plasma HIV-1 RNA > 400 copies/mL. Other endpoints were change in plasma HIV-1 RNA from baseline and percent with plasma HIV-1 RNA <400 copies/mL over time. Potential confounding variables were analyzed using the Cox proportional hazards model. Results: Compared to EFV, NVP patients had a shorter time to treatment failure (307 days vs. 589 days; p < .001), less decrease in plasma HIV-1 RNA (-0.51 log vs. -1.32 log; p < .001), and fewer patients with plasma HIV-1 RNA < 400 copies/ mL (45% vs. 51%; p < .001). Significant factors for failure were baseline CD4 count (per 100 cell increase) or viral load (per log increase), treatment center, and year of entry (p < .05 for all comparisons). Race, gender, and background nucleoside use were insignificant factors. Multivariate analysis that included significant factors for failure demonstrated improved relative hazard with EFV compared to NVP (odds ratio = 0.50, p < .001). Conclusion: EFV-containing antiretroviral regimens were associated with superior clinical outcome, as measured by time to treatment failure. Results are commensurate with other large cohort studies comparing EFV and NVP.