To test the hypothesis that subclinical enteric infection (such as bacterial overgrowth), rice malabsorption, and growth faltering are causally linked, a field trial of low-dose, short-term, intermittent antibiotic treatment was carried out in 142 hydrogen-producing (by lactulose breath hydrogen test) Burmese village children aged 6-59 months. The children were randomly allocated treatment with metronidazole (20 mg/kg or 5 mg/kg daily), amoxycillin (25 mg/kg daily), or placebo given 1 week per month for 6 months. A cooked rice meal breath hydrogen test was done to classify the children as rice absorbers (RA) or rice malabsorbers (RM) before treatment and monthly on the day before each cycle of treatment. There were no differences between the treatment groups, so they were considered together. Factorial analysis showed that antibiotic treatment did not significantly affect the proportion of RM children. The only significant difference between antibiotic-treated and placebo-treated children's growth was in the subgroup of RM children aged 36-47 months; the antibiotic-treated children had significantly greater linear growth. In other age groups antibiotic treatment had no effect on growth.