The "Feingold diet," which eliminates artificial food colorings, has been claimed to be beneficial to hyperactive children. Previous studies have yielded equivocal results. We sought to maximize the likelihood of demonstrating behavioral effects of artificial food colorings by (1) studying only children who were already on the Feingold diet and who were reported by their parents to respond markedly to artificial food colorings, (2) attempting to exclude placebo responders, and (3) administering high dosages of coloring. The design was a double-blind crossover with randomized; 11 children maintained on the Feingold diet were challenged with food coloring and placebo (one each week). Evaluations by parents, teachers, and psychiatrists and psychological testing yielded no evidence of a food coloring effect.