OBJECTIVE The relationships between adverse events (AEs) and plasma concentrations of nevirapine (NVP) and efavirenz (EFV) were investigated as part of the large, international, randomized 2NN study. METHODS Treatment-naive, HIV-1-infected patients received NVP (once or twice daily), EFV or their combination, each in combination with lamivudine and stavudine. Blood samples were collected on day 3 and weeks 1, 2, 4, 24 and 48. Concentrations of NVP and EFV were quantitatively assessed by a validated HPLC assay. Individual Bayesian estimates of the area under the plasma concentration-time curve over 24 h (AUC24h), and minimum and maximum plasma concentrations (Cmin and Cmax) as measures for drug exposure of NVP and EFV, were generated using a previously developed population pharmacokinetic model. Pharmacokinetic parameters were compared for patients with and without central nervous system (CNS) and psychiatric AEs, hepatic events, liver enzyme elevations (LEEs) and rash. Furthermore, it was investigated whether a clear cut-off for a pharmacokinetic parameter could be identified above which the incidence of AEs was clearly increased. AEs were also related to demographic parameters and baseline characteristics. RESULTS In total, from 1077 patients, NVP (3024 samples) and EFV (1694 samples) plasma concentrations and AE data (825 observations) were available. For all patients Cmin, Cmax and AUC24h were determined. When corrected for known covariates of gender, CD4 cell count at baseline, region, hepatitis coinfection and possible interactions between these factors, no significant associations between AEs and any tested exposure parameter of NVP was observed. Also, no target Cmin value, above which patients were at increased risk for AEs, could be established. On the other hand, geographical region, hepatitis coinfection, CD4 cell count and gender were found to be significantly related with the incidence of CNS and psychiatric AEs, hepatic events, LEEs and rash during the treatment with NVP. The occurrence of elevated liver enzymes during the first 6 weeks in the EFV-containing arm was significantly (P = 0.036) correlated to the exposure of EFV (Cmin). Only hepatitis coinfection impacted on LEEs during the first 6 weeks of treatment. With an EFV Cmin above 2.18 mg/l during the induction phase, patients were 4.4 (range 1.3-15.5) times more at risk for elevated liver enzymes. No other correlations between AEs and EFV pharmacokinetics or patient characteristics could be identified. CONCLUSIONS Pharmacokinetic parameters of NVP did not have a relationship to AEs in the 2NN trial when corrected for known covariates. The value of periodical drug monitoring of NVP as a way to prevent toxicity is therefore limited. Treating physicians should instead focus on factors that are more predictive of AEs (gender, CD4 count and hepatitis coinfection). High EFV Cmin levels resulted in elevated liver enzyme values during the first 6 weeks of treatment. Regular measurement of EFV levels and liver enzymes at the start of therapy may therefore be advised.