ObjectiveTo clarify the relative efficacy of nifedipine and beta-agonists for tocolysis.Data SourcesThe literature was searched in the following databases: MEDLINE 1965–1998, Embase 1988–1998, Current Contents 1997–1998, and the Cochrane Database for 1998. We also sought unpublished trials and abstracts submitted to major international congresses. Search terms were: “tocolysis,” “nifedipine,” “calcium channel blocker,” “ritodrine,” “terbutaline,” and “salbutamol.”Methods of Study SelectionRandomized controlled trials comparing tocolysis with nifedipine and beta-adrenergic agonists during preterm labor were reviewed. In cases with postrandomization exclusions, authors were contacted to obtain intent-to-treat results and to avoid analytical bias. We identified 11 published and two unpublished randomized trials.Tabulation, Integration, and ResultsData were extracted by two reviewers and analyzed by a blinded biostatistician with RevMan 3.1 software from the Cochrane Collaboration. We analyzed nine relevant randomized controlled trials that included 679 patients. Meta-analysis showed that nifedipine was more effective than the beta-agonists in delaying delivery at least 48 hours [odds ratio (OR) 1.52, 95% confidence interval (CI) 1.03, 2.24], or over 34 weeks (OR 1.87, 95% CI 1.11, 3.15). The agents did not differ as to the incidence of deliveries after 37 weeks (OR 1.29, 95% CI 0.85, 1.96) or the neonatal mortality rate (OR 1.51, 95% CI 0.63, 3.65). Treatment with nifedipine was interrupted significantly less often because of side effects (OR 0.12, 95% CI 0.05, 0.29) and led to better neonatal outcomes (fewer infants with respiratory distress syndrome: OR 0.57, 95% CI 0.37, 0.89) or transferred to neonatal intensive care units (OR 0.65, 95% CI 0.43, 0.97).ConclusionWith respect to neonatal outcome, nifedipine appears to be more effective than beta-agonists for tocolysis and should be considered for use as a first-line tocolytic agent.