Spontaneous preterm labour and delivery accounts for approximately one-third of preterm births, which is the predominant cause of perinatal mortality and morbidity. This review aims to evaluate the evidence on the benefits and harms of five classes of tocolytic therapy, namely: betamimetics, calcium channel blockers, magnesium, non-steroidal anti-inflammatory agents, and atosiban. We performed a systematic review of the effectiveness of tocolytics to stop uterine contractions (first-line therapy). Reports of randomised controlled trials from searches of MEDLINE, bibliographies of review articles, Cochrane Collaboration and its Pregnancy and Childbirth Review Group between 1966 and 2003 were identified, using the search terms "randomised controlled trial" (RCT), "preterm labor", "tocolysis", "betamimetics", "ritodrine", "prostaglandin synthetase inhibitors", "indomethacin", "calcium channel blockers", "nifedipine", "oxytocin receptor blockers", "atosiban", and "magnesium sulphate". Studies on women with preterm labour comparing the effects of a tocolytic with a placebo or no treatment that met our inclusion criteria, were included. To our knowledge, the trials were conducted mainly before 1999 and there were no placebo-controlled trials after that. Of the 86 articles identified and evaluated, 14 first-line studies met more stringent requirements for meta-analyses. Tocolytics were associated with significant decreases in the odds of delivery within 24 hours (odds-ratio [OR] 0.54, 95 percent confidence interval [CI] 0.32-0.91) and 48 hours (OR 0.47, 95 percent CI 0.30-0.75). These effects were significant for beta-agonists, atosiban and indomethacin, but not magnesium sulphate. Maternal side-effects significantly associated with betamimetics were pulmonary oedema, cardiac arrhthymias and hypokalaemia. Although calcium antagonists have not been evaluated against placebo, comparative trials with beta-agonists have shown more favourable neonatal outcomes and better prolongation of gestation. In conclusion, the management of threatened preterm labour with first-line tocolytic therapy can prolong gestation. However, the time gained in-utero need to be optimised. There is no clear first-line tocolytic agent. The use of tocolytic agents should be individualised and based on maternal condition, potential side-effects and gestational age.