Abstract Background Follow-up for 28–42 days is recommended by the World Health Organization to assess antimalarial drug efficacy for nonpregnant populations. This study aimed to determine the optimal duration for pregnant women, as no specific guidance currently exists. Methods The distributions of time to recrudescence (treatment failure), confirmed by polymerase chain reaction genotyping for different antimalarial drugs in pregnancy, were analyzed by accelerated failure time models using secondary data on microscopically confirmed recurrent falciparum malaria collected in prospective studies on the Thailand–Myanmar border between 1994 and 2010. Results Of 946 paired isolates from 703 women, the median duration of follow-up for each genotyped recurrence (interquartile range) was 129 (83–174) days, with 429 polymerase chain reaction–confirmed recrudescent. Five different treatments were evaluated, and 382 Plasmodium falciparum recrudescences were identified as eligible. With log-logistic models adjusted for baseline parasitemia, the predicted cumulative proportions of all the recrudescences that were detected by 28 days were 70% (95% confidence interval [CI], 65%–74%) for quinine monotherapy (n = 295), 66% (95% CI, 53%–76%) for artesunate monotherapy (n = 43), 62% (95% CI, 42%–79%) for artemether–lumefantrine (AL; n = 19), 46% (95% CI, 26%–67%) for artesunate with clindamycin (n = 19), and 34% (95% CI, 11%–67%) for dihydroartemisinin–piperaquine (DP; n = 6). Corresponding figures by day 42 were 89% (95% CI, 77%–95%) for AL and 71% (95% CI, 38%–91%) for DP. Follow-up for 63 days was predicted to detect ≥95% of all recrudescence, except for DP. Conclusions In low-transmission settings, antimalarial drug efficacy assessments in pregnancy require longer follow-up than for nonpregnant populations.