OBJECTIVES: Preterm birth (PTB) is a costly public health problem that causes significant neonatal morbidity and mortality. To determine cost-effectiveness of vaginal progesterone (VP) gel in the prevention of preterm birth (PTB), we developed a decision analytic model using data from the PREGNANT trial.  METHODS: PREGNANT was a multi-center, international RCT in which 459 women with singleton gestations and short cervix (10-20mm by transvaginal ultrasound) were randomized to daily VP 8% gel (n=235) or placebo (n=224).  Patient-level trial data along with cost data from the literature were used to develop the model (TreeAge Pro 2011).  Births were categorized by gestational age at delivery:  PTB at <28 weeks, PTB at 28-31 weeks, PTB at 32-36 weeks, or full term (≥37 weeks).  Costs ($US 2011) included cervical length screening, VP gel (treatment group only), antenatal hospitalizations, cerclage, and delivery hospitalization (maternal + neonatal costs). The main outcome measure was incremental cost-effectiveness, calculated as the difference in total costs between the VP gel group and the placebo group, divided by the difference in the number of PTB averted between the VP gel group and the placebo group.  A probabilistic sensitivity analysis (PSA) with 10,000 simulations was used to determine cost-effectiveness when both cost and outcome data were varied within defined limits.  RESULTS: Model base case incremental savings for VP were $12,354 and an incremental benefit of VP was 0.042 PTB averted (ICER=$21,063/PTB averted).  The PSA indicated that VP gel is expected to be less costly and more effective than placebo in 79.2% of simulated cases, however, less costly but ineffective in 16.8% of simulated cases. CONCLUSIONS:   VP gel is cost-effective in the prevention of PTB in women with short cervix as compared to placebo in most cases. Results inform ongoing clinical controversies regarding the value of VP as a preventive modality for PTB.