A controlled randomized trial was conducted in 40 infants (6-18 months old) with persistent diarrhoea (greater than 2 weeks' duration) to evaluate the effect of oral zinc supplementation. After completion of rehydration, 20 infants in group A received oral zinc sulphate (20 mg elemental zinc twice daily) and an equal number in group B were given a placebo (glucose). Each child was given oral nalidixic acid and a similar milk-free feeding schedule. Both the groups were comparable with respect to various initial characteristics including nutrition, diarrhoeal disease, serum alkaline phosphatase and serum and rectal mucosal zinc content. During therapy, all the assessed parameters of zinc status (serum alkaline phosphatase and serum and rectal zinc) recorded significant elevation and reduction in groups A and B, respectively. At recovery, the zinc status of group A was significantly higher than that of group B. The diarrhoeal duration and frequency in the zinc-supplemented group were lower but the differences were not statistically significant (p = 0.078 and p = 0.076, respectively). Weight gain in both groups was comparable. It is concluded that in persistent diarrhoea there is depletion of zinc with the progression of disease and oral zinc administration can improve the zinc status. The possible anti-diarrhoeal effect of zinc, however, merits further study.