Lamivudine population pharmacokinetics were investigated by using nonlinear mixed-effect modelling (NONMEM) analysis of data from 394 human immunodeficiency virus (HIV)-infected patients treated with lamivudine (150 to 300 mg every 12 h) in two large, phase III clinical efficacy-safety trials, NUCA3001 and NUCA3002. Analyses of 1,477 serum lamivudine concentration determinations showed that population estimates for lamivudine oral clearance (CL/F; 25.1 liters/h) and volume of distribution (V/F; 128 liters) were similar to values previously reported for HIV-infected patients in phase I pharmacokinetic studies. Lamivudine CL/F was significantly influenced by the covariates creatinine clearance and weight and not affected by age, Centers for Disease Control and Prevention (CDC) classification, CD4(+) cell count, HIV type 1 (HIV-1) RNA PCR, or gender and race when CL/F was corrected for differences in patient weight. The population estimate for lamivudine V/F was not significantly influenced by the covariates gender, race, age, weight, renal function, HIV-1 RNA PCR, or CDC classification and CD4(+) cell count when creatinine clearance was included with CL/F in the model. Lamivudine disposition was significantly influenced by renal function. However, as only three patients had an estimated creatinine clearance of <60 ml/min, dosage adjustments for patients with impaired renal function should not be determined based on the population parameters derived in this analysis.