Preterm birth is the leading cause of neonatal morbidity and mortality, and previous preterm birth is one of the strongest risk factors for preterm birth. National and international obstetrical societies have different recommendations regarding progesterone formulation for the prevention of recurrent preterm birth. This study aimed to determine whether vaginal progesterone is superior to 17-hydroxyprogesterone caproate in the prevention of recurrent preterm birth in patients with singleton pregnancies who had a previous spontaneous preterm birth. This was an open-label multicenter pragmatic randomized controlled trial at 5 US centers of patients with singleton pregnancies at <24 weeks of gestation who had a previous spontaneous preterm birth randomized 1:1 to either 200 mg vaginal progesterone suppository nightly or 250 mg intramuscular 17-hydroxyprogesterone caproate weekly from 16 to 36 weeks of gestation. Based on the estimated recurrent preterm birth rate of 36% with 17-hydroxyprogesterone caproate, 95 participants were needed in each arm to detect a 50% reduction in preterm birth rate with vaginal progesterone, with 80% power and 2-sided alpha of 0.05. The primary outcome was preterm birth at <37 weeks of gestation. Prespecified secondary outcomes included preterm birth at <34 and <28 weeks of gestation, mean gestational age at delivery, neonatal morbidity and mortality, and measures of adherence. Analysis was by intention to treat. The chi-square test and Student t test were used as appropriate. P<.05 was considered significant. Overall, 205 participants were randomized; 94 participants in the vaginal progesterone group and 94 participants in 17-hydroxyprogesterone caproate group were included. Although gestational age at enrollment was similar, those assigned to vaginal progesterone initiated therapy earlier (16.9±1.4 vs 17.8±2.5 weeks; P=.001). Overall continuation of assigned formulation until delivery was similar (73% vs 69%; P=.61). There was no significant difference in preterm birth at <37 (31% vs 38%; P=.28; relative risk, 0.81 [95% confidence interval, 0.54-1.20]), <34 (9.6% vs 14.9%; P=.26; relative risk, 0.64 [95% confidence interval, 0.29-1.41]), or <28 (1.1% vs 4.3%; P=.37; relative risk, 0.25 [95% confidence interval, 0.03-2.20]) weeks of gestation. Participants in the vaginal progesterone group had a later mean gestational age at delivery than participants in the 17-hydroxyprogesterone caproate group (37.36±2.72 vs 36.34±4.10 weeks; mean difference, 1.02 [95% confidence interval, 0.01-2.01]; P=.047). Vaginal progesterone did not reduce the risk of recurrent preterm birth by 50% compared with 17-OHPC; however, vaginal progesterone may lead to increased latency to delivery. This trial was underpowered to detect a smaller, but still clinically significant, difference in the efficacy of preterm birth prevention. Patient factors that impact adherence and ability to obtain medication in a timely fashion should be included in counseling on progesterone selection.