Clinical performance measurement is an increasingly important way for public and private purchasers to compare the value of health services provided by competing health delivery systems. The widespread use of performance measures has increased the demand for development of new measures that cover previously unevaluated aspects of care. Four steps required to develop a clinical performance measure that is suitable for making comparisons among health delivery systems are discussed: (1) choosing clinical areas to measure, (2) selecting performance indicators within each area, (3) designing specifications for consistent implementation measures, and (4) evaluating the scientific strength of a measure. The application of these steps to developing measures of quality for hypertension is provided, emphasizing a measure of adequacy of control of blood pressure. Developing useful clinical performance measures requires careful attention to methodologic issues. The steps outlined in this paper should enhance the quality of future measurement development.