In a double-blind trial lasting 4 months in 42 mentally handicapped patients, the effect of lithium on aggression was assessed in comparison with placebo. In the lithium-treated group, 73% of patients showed a reduction in aggression during treatment. There were significant differences in mean weekly aggression scores and in the frequency of aggressive episodes between the lithium and placebo groups. Side-effects were noted in 36% of the lithium group (and 20% of the placebo group), but were mainly transitory. There were no episodes of toxicity, and no patients had to be withdrawn from the trial. Lithium appears to be worth a 2-month trial in such patients, where repeated aggression has not been relieved by more appropriate placement, occupation or company.