A program for the recruitment of control subject is examined by considering the yield of healthy subjects as a function of the initial screening interview. All subjects who passed the screen also completed a diagnostic evaluation for lifetime history of mental disorders. Version 1 of the screen focused on excluding individuals who were currently mentally ill. Approximately 20% met criteria for Never Mentally Ill (NMI); 7% had one episode of a minor mental disorder (MMD); 43% were currently not mentally ill, but had a serious history of mental illness (CNMI); and 30% were currently mentally ill (CMI). Revision of the screen to exclude individuals with any history of treatment for mental illness significantly improved the proportion of NMI subjects (29%), and there was a trend for a decreased proportion of CMI subjects (25%). Additional revision of the screen to exclude individuals with a history of selected untreated mental disorders resulted in a significantly greater proportion of NMI subjects (50%), and a significantly lower proportion of CMI (17%) subjects. The results of this investigation demonstrate that the cost-effectiveness of the initial screen can be greatly improved by expanding the evaluation to include selected untreated as well as treated mental disorders.