BACKGROUND: Enuresis (bedwetting) is a socially disruptive and stressful condition which affects around 15-20% of five year olds, and up to 2% of young adults. OBJECTIVES: To assess the effects of tricyclic and related drugs on nocturnal enuresis in children, and to compare them with other interventions. SEARCH STRATEGY: We searched the Cochrane Incontinence Group trials register (December 2002) and the reference lists of relevant articles including two previously published versions of this review. Date of the most recent searches: December 2002. SELECTION CRITERIA: All randomised and quasi-randomised trials of tricyclics or related drugs for nocturnal enuresis in children were included in the review. Comparison interventions included placebo, other drugs, alarms, behavioural methods or complementary/miscellaneous interventions. Trials focused solely on daytime wetting were excluded. DATA COLLECTION AND ANALYSIS: Two reviewers independently assessed the quality of the eligible trials, and extracted data. MAIN RESULTS: Fifty four randomised trials met the inclusion criteria, involving 3379 children. The quality of many of the trials was poor. Most comparisons or outcomes were addressed only by single trials. Treatment with most tricyclic drugs (such as imipramine, amitriptyline, viloxazine, nortriptyline, clomipramine and desipramine) was associated with a reduction of about one wet night per week while on treatment (eg imipramine compared with placebo, weighted mean difference (WMD) -1.19, 95% CI -1.56 to -0.82). The exception was mianserin, where results from one small trial did not reach statistical significance. About a fifth of the children became dry while on treatment (relative risk for failure (RR) 0.77, 95% CI 0.72 to 0.83), but this effect was not sustained after treatment stopped (eg imipramine versus placebo, RR 0.98, 95% CI 0.95 to 1.03). There was not enough information to assess the relative performance of one tricyclic against another, except that imipramine was better than mianserin. The evidence comparing desmopressin with tricyclics was unreliable or conflicting, but in one small trial all the children failed or relapsed after stopping active treatment with either drug.The evidence comparing tricyclics with alarms was also unreliable or conflicting during treatment. In one small trial all the children failed or relapsed after tricyclics stopped, compared with about half after alarms. This result was compatible with the results in the Cochrane review of alarm treatment, which found that about half the children remained dry after alarm treatment was finished. There was a little evidence from single trials to suggest that imipramine might be better than a simple reward system with star charts during treatment; worse than a complex intervention involving education, counseling, waking and retention control training; better than a restricted diet; and worse than hypnosis. However, these results need to be confirmed by further research. REVIEWER'S CONCLUSIONS: Although tricyclics and desmopressin are effective in reducing the number of wet nights while taking the drugs, most children relapse after stopping active treatment. In contrast, only half the children relapse after alarm treatment. Parents should be warned of the potentially serious adverse effects of tricyclic overdose when choosing treatment. Further research is needed into comparisons between drug and behavioural or complementary treatments, and should include relapse rates after treatment is finished.