At immunization contact, 165 infants 2.5 mo old were randomly assigned to receive either 15 mg vitamin A (retinyl palmitate) or placebo. Three doses were given at monthly intervals with each diphtheria, pertussis, tetanus and oral polio (DPT/OPV) immunization dose. The diarrhea and acute respiratory infection (ARI) morbidity was similar in the vitamin A and placebo groups. However, the duration (days per child-year, mean +/- SD) of ARI was less in the vitamin A group compared with placebo group (27.6 +/- 17.1 vs. 40.8 +/- 22.7; P = 0.005). Fasting retinol concentrations were measured at entry and in 61 infants, the relative dose response (RDR) test was done 1 mo after the third dose of vitamin A. Eighty-five percent of the infants had serum retinol concentration < 0.70 mol/L at entry. After 3 mo the serum retinol levels improved significantly in both groups, and in the vitamin A-supplemented group the serum retinol concentration was significantly better than that in the placebo group (P= 0.02). However, 61% of the infants remained deficient despite vitamin A supplementation. Among vitamin A-supplemented infants only, diarrhea and ARI morbidity during the 3-mo period were compared in children with normal versus children with abnormal RDR at the end of the supplementation period. The ARI episodes were more frequent in the supplemented infants who remained vitamin A deficient at the end of the 3 mo (P = 0.027). Also, the cumulative duration (days, mean +/- SD) of fever and cough was 5.0 +/- 2.8 in the normal versus 11.2 +/- 6.0 in the deficient group (P = 0.04). The results of this study suggest that a large proportion of infants remain vitamin A deficient even after large dose vitamin A supplementation because of frequent respiratory infections, particularly those accompanied by fever.