In a double-masked, placebo-controlled, clinical trial in Indonesia, 88 preschool children between the ages of 36 and 72 months with Bitot's spots were randomly assigned to 200,000 IU of oral vitamin A or placebo and followed up for five weeks. Baseline and follow-up serum vitamin A levels were obtained. Of the 45 children receiving vitamin A, 33 (73.3%) showed complete cure and disappearance of Bitot's spots, six (13.3%) had disappearance of some but not all Bitot's spots, and six (13.3%) were unresponsive to treatment. The nonresponsive children were older, all male, and had higher initial mean serum vitamin A levels, 23.0 micrograms/dl, compared to 15.9 micrograms/dl in the cured group (P = .017). These data suggest that normal vitamin A status may be found in approximately 13% of children with Bitot's spots before vitamin A intervention and that one fourth of those who had Bitot's spots originally will not be cured of all Bitot's spots after intervention. These are important factors to consider when using Bitot's spots in prevalence surveys as a clinical sign of vitamin A deficiency.