In bone densitometry, the precision of the instrument, the number of measurements and the time-points of the measurements are important criteria for monitoring bone mass changes. The most appropriate follow-up procedure can be determined by numerical comparison of various combinations of these three criteria. This can be done by computing the confidence interval of changes in bone mass. We developed a model to estimate the length of a confidence interval for the observed changes in individual patients. With specific instrument precision, a specified number of measurements and, assuming a linear rate of bone mass changes, the best estimate of the actual changes in bone mass is obtained by measurements at the end of an observation period. With the current precision of bone densitometers, follow-up of patients with yearly duplicate measurements is recommended. A shorter scan time interval offers no additional information unless very rapid bone loss is expected.