The health care industry has been faced with increasing pressure to control costs. Health care costs have increased substantially more rapidly than general inflation rates. At the same time, health care facilities face price competition for services because insurance companies and government-funded health programs are limiting opportunities for cost reimbursement. To control costs, one must first relate the costs of providing services to the volume of activity. The first step is often to estimate a cost model, TC = F + VX, where X refers to the volume of activity. Examples of activity bases include patient days to estimate nursing staff costs or number of tests to estimate costs in a laboratory. Although it may appear simple to estimate the relation TC = F + VX, analysts often find a lack of good data to make the estimates. For example, the cost of medical supplies shown in the accounting records is often the cost of purchases, not the cost of supplies used. Consequently, large purchases in one month followed by no purchases in the next month make these costs appear to behave in unrealistic ways. Although recent pressures on health care facilities to reduce costs have increased the incentives for administrators and doctors to improve record keeping, our research indicates the information needed to control costs is lacking in many health care organizations. For example, hospitals often keep track of charges to patients but not the costs of the items being charged. How would information that makes it possible to estimate the equation TC = F + VX help health care managers control costs?