The first step in developing this information base is to more clearly determine the actual relationships between the criminal justice and mental health systems. After these relationships have been clarified, issues concerned with the selection, training, and program needs of jail staff; the goals and effects of jail mental health services; and the processes of effective program development and implementation can be more appropriately addressed. A general strategy that is especially adapted both to the questions of the overall relationships between the jails and mental health services and to the effects of mental health treatment programs in jails is that of cohort studies. This means that large groups of persons are selected for study and their paths through the criminal justice and mental health systems are tracked for many years. These cohort studies should be structured for the use of control groups for comparative purposes. To provide a climate in which the interests of individual researchers and the needs of agencies or regions can be merged, Federal agencies should specify generic problem areas as guidelines, leaving the individual practitioners and researchers to pose specific questions. A total of 19 references are provided.