A total of 24 subjects, 16 males and 8 females, average age 29.4 +/- 6.5 y, were randomly divided into two groups. Group A received 400 micrograms/d of lithium orally, in tablets composed of a naturally lithium-rich brewer's yeast, for 4 wk. Group B was given normal, lithium-free brewer's yeast as a placebo. All the subjects of the study were former drug users (mostly heroin and crystal methamphetamine). Some of the subjects were violent offenders or had a history of domestic violence. The subjects completed weekly self-administered mood test questionnaires, which contained 29 items covering parameters measuring mental and physical activity, ability to think and work, mood, and emotionality. In the lithium group, the total mood test scores increased steadily and significantly during the period of supplementation. The 29 items were furthermore placed into three subcategories reflecting happiness, friendliness, and energy, as well as their negative counterparts. In Group A, the scores increased consistently for all subcategories until wk 4 and remained essentially the same in wk 5. In Group B, the combined mood test scores showed no consistent changes during the same period. The only positive change in some members of Group B occurred during wk 1 and was attributed to a placebo effect. In Group B, the placebo effect was noticeable for the subcategories of energy and friendliness; the happiness scores declined during the entire period of observation. Based on these results and the analysis of voluntary written comments of study participants, it is concluded that lithium at the dosages chosen had a mood-improving and -stabilizing effect.