The efficacy of single-dose antibiotics for treatment of symptomatic urinary tract infections in women was compared to the efficacy of conventional therapy in 25 controlled studies. The combined results of these studies indicate that single-dose antibiotics are less effective than conventional therapy at 3-14 days post-treatment (odds ratio of 0.5, 95 per cent confidence interval (C.I.) 0.4-0.7), and at 4-6 weeks (odds ratio of 0.4, 95 per cent C.I. 0.3-0.6). Single-dose therapy did not perform better in patients with bacterial isolates sensitive to the drug used. The cure rate achieved by single-dose amoxicillin was significantly lower than that afforded by conventional therapy. The equal efficacy of single-dose and conventional trimethoprim/sulphamethoxazole treatment at 3-14 days post-treatment could not be ruled out with a 95 per cent certainty. However, conventional treatment achieved cure rates significantly higher than single-dose therapy at 4-6 weeks (odds ratio 0.48, 95 per cent C.I. 0.3-0.8). Side-effects were lower among patients given a single dose (odds ratio 0.5, 95 per cent C.I. 0.4-0.6). Although single-dose trimethoprim/sulphamethoxazole is less effective than a course of treatment, it causes fewer side-effects: the decision of whether to use a single-dose has to be decided by cost-benefit analysis in any specific health-care system.