To determine whether zinc deficiency might be involved in the failure to thrive observed in undernourished Vietnamese children, we assessed growth, incidence of infections, and circulating insulin-like growth factor I (IGF-I) concentrations in a double-blind study of zinc supplementation. Growth-retarded children (n = 146) aged 4-36 mo were paired according to age, sex, commune, Z scores for weight (WAZ) and for height (HAZ), and number of siblings, and were randomly assigned to receive either 153 mu mol (10 mg) Zn/d or a placebo for 5 mo. Weight, height, and episodes of infection were recorded each month and plasma IGF-I was measured 1 and 5 mo after the start of zinc supplementation. Multiple-linear-regression analysis for paired data showed that zinc supplementation increased weight (+0.5 +/- 0.1 kg; P < 0.001) and height (+1.5 +/- 0.2 cm; P < 0.001) after 5 mo compared with placebo treatment. The relative risk of infectious episodes in the zinc-treated subjects was reduced S-fold for diarrhea (P = 0.012) and 2.5-fold for respiratory infections (P = 0.057). The probability of having at least two episodes of diarrhea or respiratory infection was 2.9- and 3.2-fold lower, respectively, in zinc-treated than in placebo-treated children. Plasma IGF-I concentrations increased in zinc-treated subjects between 1 and 5 mo (P = 0.018), whereas they did not change in placebo-treated subjects (P = 0.584). After 1 mo, IGF-I concentration ((x) over bar +/- SD) in zinc-treated subjects was 2.8 +/- 0.3 nmol/L compared with 1.9 +/- 0.2 nmol/L in placebo-treated subjects (P = 0.021). After 5 mo, the values were 3.4 +/- 0.5 nmol/L (zinc-treated) and 2.0 +/- 0.3 nmol/L (placebo-treated; P = 0.044). Our study suggests that zinc deficiency may limit growth in nutritionally deprived children. Because the increase in growth velocity resulting from zinc supplementation was associated with increased plasma IGF-I concentrations, we suggest that the growth-stimulating effect of zinc might be mediated through changes in circulating IGF-I.