H EALTH departments are constantly coping with the problems of insufficient money and people to carry on their present programs; at the same time, the rapidly growinig population is making demands for new and increased health services. Facts concerning past and present services to the current programs are sorely needed as a means of evaluating past services, redesigning current operations, and determining priorities for the future. How can we obtain these facts in a simple manner? Most public health workers are already overbuirdened witlh recordkeeping, and we need to be very cautious about adding to this load. However, many of the present recordkeeping systems are designed to provide specific information regarding a certain program without thought of relating it to other informationi concerning the same program or to other programs. The scope of public health services is continually broadening, and more and more disciplines are represented on the staffs of health departments. The result is that no one program is conducted solely by one organizational uniit of the department, but rather through the cooperative effort of many disciplines from several program ulnits. For example, usually the tuberculosis services of a health department include the work of not only the tuberculosis division but also other divisions such as nursing, nutrition, education, social service, and environmental health. This interdependence of disciplines is increasing rather than decreasing, and as it inlereases so will the need increase for interrelated statistics. To collect service statistics from several units concerning a single program is both a necessity and a problem. For some time, statisticians of the Michigan Department of Health had been disturbed by the lack of service statistics compiled by program and had recognized a, need for a system to provide these data on a departmentwide basis regardless of organizational units. A new system began in the Michigan Department of Health the day the chief of the nutrition section came to the statistical methods section for help with the reporting system for her section. She commented that the daily reports were time consuming to tabulate, and once counted had little meaning for they could not be related to each other and, consequently, were of little value for program planning or evaluation. The nutrition section seemed a logical unit for a pilot study in the use of a new system. There were only five members on the staff, and they Miss Duxbury is chief of the statistical methods section in the division of disease control, records, and statistics, of the Michigan Department of Health, Lansing.