A comparison of the long-term effects of four kinds of psychotropic medications. Mood disorders are probably the leading diagnostic category in general practice, and the effects of antidepressant medication are well documented in the literature. Long-term clinical observations of the effect of psychoactive medications have not been so well reported. This study, therefore, investigated the long-term effects of four major categories of psychoactive medication on patients attending an inpatient psychiatric facility: 1) tricyclic antidepressants (TCAs); 2) serotonin selective reuptake inhibitors (SSRIs); 3) lithium salts; and 4) benzodiazepines. It is a retrospective study, and information on 1205 patients was collected. The results are reported here in order to provide more information about the long-term effects of mood stabilizers, antidepressants and anxiolytic medication in common practice. The average patient, aged 65 years, was female, and in regular psychiatric care for 6.5 years. A considerable proportion of the sample (11.2%) had been treated for other psychiatric disorders and half of these patients had received three or more different psychotropic drugs in the past. Patients were treated with one or more drugs in each group during their stay at the institution and they were mainly treated with only one of these groups at any given time. A substantial proportion of patients (50.9%) had been exposed to three different groups of medication over their course of psychiatric treatment. Lithium salts were the most widely used medication in the four treatment groups. Antidepressants were the second most frequently used group of drugs. Our results show that the two groups of drugs have comparable effects in terms of morbidity. However, lithium was used more as a prophylactic measure than as an acute-phase remedy and was shown to be superior to the other three groups of drugs in preventing relapse into major depression. These findings are consistent with the concept of mood stabilizers as an effective long-term treatment for depressive episodes, although the number of patients with a past history of antidepressant treatment precludes definitive conclusions. The data indicate that psychiatrists do not usually evaluate the long-term effects of their own prescribed medication. Such effects may have considerable clinical relevance and warrant further exploration.