In scientific observations on the utilization of drugs under routine treatment conditions in patients with affective disorders, one of the main problems is the establishment of criteria for measuring treatment effectiveness. As the mortality risk of such patients is considerable, the mortality rate is an important outcome criterion for determining the long-term effectiveness of medication administered in routine treatment settings. Mortality studies are typical phase IV studies; the researcher does not interact with the treatment procedure, he only observes the results of treatment with respect to death rates. As lithium treatment is of increasing importance and interest as a major prophylactic agent in recurrent affective disorders and is by its nature a long-term treatment, such a therapy provides an excellent field for observations of effectiveness under long-term treatment conditions. Mortality studies can take into account various factors, such as type of treatment setting, treatment regimen, drop-out analysis, and comparison between times on and off treatment. Results from several mortality studies on lithium-treated patients are reported in order to demonstrate the methodology and outcome of long-term observation of drug treatment. Mortality studies conducted within the framework of phase IV studies and particularly within the concept of pharmacoeconomics can help to demonstrate the effectiveness of long-term treatment, and are an important methodological adjunct to controlled clinical trials.