The authors report the case of a patient who developed hyperammonemia and coma during therapy with valproic acid for affective disorder. Onset of the coma was gradual and initially interpreted as a therapeutic reduction in the patient's anxiety. In a psychiatric setting, treatment of hyperammonemia may be delayed if a patient's increasing lethargy is interpreted as a therapeutic response. Staff may need to be educated about the potential for hyperammonemia, and patients whose tolerance for valproic acid is unknown may need to be monitored for liver function and blood levels of urea and ammonia.