The objective of this study was to determine how effective angiotensin-converting enzymes (ACEs) must be in preventing diabetic nephropathy to warrant routine administration to insulin-dependent diabetic patients. A Markov model was used to compare three strategies designed to prevent the development of end-stage renal disease in insulin-dependent diabetic patients. Strategy I, screening for microalbuminuria and treatment of incipient nephropathy as currently recommended, was compared with strategy II, a protocol in which patients were routinely administered an ACE inhibitor 5 years after diagnosis of diabetes, and strategy III, in which patients at high risk for nephropathy were routinely treated and low-risk patients followed a protocol in which patients were treated with an ACE inhibitor if they developed hypertension and/or macroproteinuria. The model predicted that strategy II would produce as many quality-adjusted life-years as strategy I at nearly the same cost if routine drug therapy reduced the rate of development of microalbuminuria by 26% in all patients. Strategy III produced as many quality-adjusted life-years at less cost than strategy I if a high-risk cohort could be identified with a rate of developing microalbuminuria at four times the rate of low-risk patients and if drug therapy reduced the rate of developing microalbuminuria in this high-risk group by 20%. In conclusion, routine ACE inhibitor therapy could prove to be cost-effective, especially if high-risk individuals could be identified. A prospective trial examining this goal should be considered.