OBJECTIVE: To measure nevirapine elimination in African adults undertaking a structured treatment interruption (STI) in the DART trial. DESIGN: Cohort (16 women, 5 men; median weight 61 kg) within a randomized trial of management strategies. METHODS: Plasma nevirapine was measured by validated high performance liquid chromatography at 0,1,2,3 and 4 weeks after stopping the drug in a subset of patients undertaking an STI. All patients continued lamivudine plus zidovudine/stavudine for a further 7 days. RESULTS: Two patients with no or low plasma nevirapine concentration at baseline were excluded. Geometric mean plasma concentration when nevirapine was stopped in the remaining 19 patients was 6421 ng/ml (range, 3724-9473). Nevirapine was detected in 15/18 (83%) patients at 1 week, and 5/19 (26%) patients at 2 weeks but was not found any samples collected after 2 weeks. Only one patient had > 100 ng/ml (limit of quantification) at 2 weeks (415 ng/ml, female). The median times to reach thresholds of 200, 100 and 20 ng/ml (limit of detection) were estimated to be 7.6 [interquartile range (IQR), 7.0-10.1], 9.3 (IQR, 8.7-13.0) and 13.2 (IQR, 12.3-18.4) days, respectively, with 3/19 (16%) and 14/19 (74%) estimated to have reached < 20 ng/ml by 7 and 14 days, respectively. CONCLUSION: Although elimination of nevirapine was faster than previously published after a single dose, the data suggest that an additional staggered period of 7-10 days with dual nucleotide reverse transcriptase inhibitor cover is necessary for African patients discontinuing nevirapine.