Nonsteroidal antiinflammatory drugs (NSAIDs) are increasingly being used during pregnancy to treat a variety of conditions. An evaluation of the risk of premature closure of the ductus arteriosus is useful in determining the safety of NSAIDs at different stages of pregnancy. To determine whether NSAID use during the third trimester of pregnancy is associated with an increased risk of premature constriction of the ductus arteriosus. A systematic review was conducted of MEDLINE (1966-2004), Embase (1980-2004), and the Cochrane Database of Systematic Reviews (1991-2004). Summary estimates of the odds ratios, comparing ductal outcomes in exposed and unexposed fetuses, and their 95% confidence intervals were calculated assuming a random effects model. Based on 217 patients exposed to indomethacin and 221 to placebo, the risk of ductal closure was 15-fold higher in the group of women exposed to NSAIDs compared with those receiving either placebo or other NSAIDs (8 studies; OR = 15.04, 95% CI 3.29 to 68.68). There was no significant increased risk of ductal closure in the infants of women treated with indomethacin compared with those receiving other drugs (4 studies; OR = 2.12, 95% CI 0.48 to 9.25). Similar results were found when calculating rate differences. Short-term use of NSAIDs in late pregnancy is associated with a significant increase in the risk of premature ductal closure.