An open, randomised comparison of 2 or 3 days of oral ofloxacin (10 mg/kg/day) for uncomplicated typhoid fever was conducted in 235 Vietnamese children. Multi-drug-resistant Salmonella typhi was isolated from 182/202 (90%) children and 5/166 (3%) tested isolates were nalidixic acid-resistant (Na(R)). Eighty-nine of 116 children randomised to 2 days and 107/119 randomised to 3 days were blood culture-positive and eligible for analysis. There were 12 (13.5%) failures in the 2-day group (six clinical failures, four blood culture-positive post treatment, two relapses) compared with eight (7.5%) failures in the 3-day group (four clinical failures, one blood culture-positive post treatment, three relapses) (OR 1.9, 95% CI 0.7-5.5,p = 0.17). There were no significant differences in the mean (95% confidence interval) fever clearance times (h) [92 (82-102) vs 101 (93-110), p = 0.18] or duration of hospitalisation (d) [7.6 (7.2-8.1) vs 8.0 (7.6-8.4), p = 0.19] between the two groups. There was one failure in the four eligible children infected with an Na(R) isolate of S. typhi. No adverse events were attributable to the ofloxacin. These results extend previous observations on the efficacy of short courses of ofloxacin for children with uncomplicated multi-drug-resistant typhoid fever.