Objective: To assess the impact of HIV-1 protease and reverse transcriptase (RT) mutations, and pharmacokinetic parameters on virological responses to nelfinavir (NFV)-containing highly active antiretroviral therapy. Design: Naive or antiretroviral-experienced HIV-1-infected subjects were included in a non-randomized, observational cohort study and received two nucleoside RT inhibitors + NFV (750 mg three times per day or 1250 mg twice per day). Virologic success was defined as a virus load < 50 copies/ml for > 6 months. Methods: RT and protease genes were sequenced at baseline and at the time of virological failure. Plasma NFV trough concentration (C-min,), maximum concentration (C-max), and AUC(0-tau) at steady-state were subjected to population pharmacokinetic analysis. Results: Patients (n = 154) enrolled between November 1998 and February 2000 started a twice per day (n = 84) or three times per day (n = 70) NFV-based regimen as first- (n = 48) or second-line therapy when protease inhibitor-naive (n = 64) or experienced (n = 42). Median follow-up duration was 16 months. Virologic failure occurred in 88 patients. No significant differences were observed between twice per day and three times per day regimens. According to multivariate analysis, NFV C-min and C-max, CD4 cell count, number of baseline RT + protease gene mutations, D67N, M184V, T215F/Y in RT, and M361 in protease, were independent factors that were significantly predictive of failure. At failure, L101, D30N, M361, V771, N88S/D or L90M protease mutations had emerged since baseline. Pharmacokinetic parameters were similar in patients with or without emergence of these neo-mutations. The more discriminating NFV C-min efficacy-threshold was estimated to be 1 mg/l. Conclusions: Our data confirm the association among individual pharmacokinetic parameters, genotype pattern and virological response to NFV-containing regimens. (C) 2002 Lippincott Williams Wilkins.