A comparison of the pharmacokinetics of oral metronidazole, after a single dose of 30 mg/kg body weight, was done in two groups of subjects: group I consisted of 10 severely malnourished children, aged 4 to 43 months; group II consisted of 10 children, aged 3 to 25 months, who were studied after nutritional rehabilitation. The biologic half-life of elimination was significantly longer (p less than 0.01) in severely malnourished children (median, 10.21 hours; range, 4.89 to 22.93 hours) than in rehabilitated children (median, 5.09 hours; range, 2.61 to 8.75 hours). Metabolic clearance of metronidazole was significantly lower in group I (p less than 0.01; median, 0.077 L/kg/hr; range, 0.033 to 0.192 L/kg/hr) than in nutritionally rehabilitated children (median, 0.166 L/kg/hr; range, 0.105 to 0.300 L/kg/hr). Volume of distribution was not different between groups I and II, although both showed higher values than the values reported for children who were not malnourished. These findings suggest that the dose of metronidazole should be reduced in malnourished children, and the therapeutic regimen should be individualized for each patient.