BACKGROUND Cholera is a major public-health problem, with children most affected. However, effective single-dose antimicrobial regimens have been identified only for adults. Our aim was to compare the efficacy of azithromycin and erythromycin regimens in the treatment of children. METHODS We did a double-blind, randomised study of 128 severely dehydrated children (age 1-15 years) with cholera, treated at one of two treatment centres in Bangladesh in 1999. Children were assigned single-dose azithromycin (20 mg/kg bodyweight, maximum individual dose 1 g; n=65) or 12.5 mg/kg erythromycin (maximum dose 500 mg; n=63) every 6 h for 3 days. Patients stayed in hospital for 5 days. We measured fluid balance every 6 h, and obtained a rectal swab or stool sample for culture daily. Our primary outcome measures were clinical success of treatment-ie, cessation of watery diarrhoea within 48 h-and bacteriological success-ie, absence of Vibrio cholerae O1 or O139 from cultures of stool or rectal swab samples after study day 2. Analysis was per protocol. RESULTS Two children in both groups withdrew from the study, and we excluded one child in the erythromycin group. Treatment was clinically successful in 48 (76%) patients who received azithromycin and 39 (65%) who received erythromycin (difference 11%, 95% CI -5 to 27, p=0.244); and bacteriologically successful in 45 (71%) and 49 (82%) patients, respectively (10%, -5 to 25, p=0.261). Patients treated with azithromycin had a shorter duration of diarrhoea (median 24 h vs 42 h; difference 12 h, 0-18 h, p=0.019) and fewer episodes of vomiting (1 vs 4; difference 1 episode, 0-3 episodes, p=0.023). CONCLUSIONS Single-dose azithromycin is as effective for treatment of cholera in children as standard erythromycin therapy, but is associated with less vomiting.