Since 1999 an ongoing international interlaboratory quality control program has analyzed antiretroviral drugs in plasma. Results of the third round of this program are presented. Quality control samples were prepared by spiking drug-free plasma with varying concentrations of the currently available protease inhibitors and the nonnucleoside reverse transcriptase inhibitors efavirenz and nevirapine. Thirty-three laboratories participated in the program and were requested to analyze the quality control samples. Results were from 30 laboratories. Of all measurements, 82% were performed within 80%-120% accuracy limits. Only 3 laboratories performed all their measurements within these limits, and 12 participants reported at least 90% of their analyses within the acceptance range. Mean accuracy for low drug concentrations was worse than for medium and high concentrations. The percentage of satisfactory measurements for the 6 laboratories that participated for the third time in the program increased from 54% in the first round to 85% in the third round. The program revealed a large variability in the laboratories' ability to measure antiretroviral drugs accurately. This variability may have important implications for therapeutic drug monitoring of these drugs and for pharmacokinetic studies. Interlaboratory testing is useful to alert laboratories to previously undetected analytical problems.