Forty children were given a diet free of artificial food dyes and other additives for 5 days. Twenty of the children had been classified as hyperactive by scores on the Conners Rating Scale and were reported to have favorable responses to stimulant medication. A diagnosis of hyperactivity had been rejected in the other 20 children. Oral challenges with large doses (100 or 150 milligrams) of a blend of FD & C approved food dyes or placebo were administered on days 4 and 5 of the experiment. The performance of the hyperactive children on paired-associate learning tests on the day they received the dye blend was impaired relative to their performance after they received the placebo, but the performance of the nonhyperactive group was not affected by the challenge with the food dye blend.