BackgroundIn vitro passage experiments are crucial to the development of antiretroviral (ARV) drugs. MethodsWe created an online database containing data from 102 published studies in which HIV-1 or HIV-2 was cultured with increasing concentrations of the FDA-approved nucleoside RT inhibitors (NRTIs), nonnucleoside RT inhibitors (NNRTIs), integrase strand transfer inhibitors (INSTIs), protease inhibitors (PIs), capsid inhibitor (CAI) lenacapavir, and nucleoside RT translocation inhibitor (NRTTI) islatravir. We summarized the mutations selected in the subset of passage experiments with NRTIs lamivudine (3TC), emtricitabine (FTC), abacavir (ABC), tenofovir (TFV), and zidovudine (AZT), NNRTIs doravirine (DOR), efavirenz (EFV), and rilpivirine (RPV), INSTIs bictegravir (BIC), cabotegravir (CAB), and dolutegravir (DTG), and PIs atazanavir (ATV), darunavir (DRV), and lopinavir (LPV). Mutations selected in vitro were compared with those selected in persons receiving the same ARV. ResultsTwenty-seven studies described 89 experiments of wildtype isolates passaged with 3TC, FTC, ABC, TFV, or AZT; sixteen studies described 89 experiments passaged with EFV, RPV, or DOR; eleven studies described 76 experiments passaged with the INSTIs BIC, CAB, or DTG; six studies described 33 experiments passaged with ATV, LPV, or DRV. With several exceptions, mutations selected in two or more experiments were among the most common mutations selected in persons receiving the same ARV. ConclusionsWe created a database of published ARV in vitro selection experiments. Mutations emerging from these experiments generally predict those observed in persons receiving the same ARV. However, there are notable differences in mutation frequencies between in vitro and in vivo settings.