To study how the risk of cardiovascular disease changes with increasing levels of urinary albumin excretion (UAE), we prospectively studied a random sample of 120 49-year-old men with a wide range of blood pressures. Based on diastolic blood pressure (DBP), the subjects were divided into normotensives (DBP < 90 mm Hg; n = 21), borderline hypertensives (DBP 90 to 94 mm Hg; n = 30), mild hypertensives (DBP 95 to 104 mm Hg; n = 45) and moderate to severe hypertensives (DBP > 105 mm Hg; n = 24). None had been previously treated for hypertension or had secondary hypertension, diabetes mellitus, or other cardiovascular diseases at baseline. Heart and kidney function and metabolic and hormonal variables were with beta-blockade, diuretics, or hydralazine. The cardiovascular morbidity during 10 years of follow-up was studied. The hypertensives were treated with beta-blockade, diuretics, or hydralazine. The cardiovascular morbidity during 10 years of follow-up was studied. The 19 subjects who developed cardiovascular disease had significantly higher baseline UAE than the group that did not (median value 16.6 mg/24 h; range 3.5 to 73, and 9.7 mg/24 h, range 0 to 308, respectively). UAE correlated to systolic blood pressure (P = .0115) and DBP (P = .031), but not to smoking behavior or serum cholesterol. The risk of cardiovascular disease was associated with UAE and smoking independently of blood pressure (P = .001 and P = .015, respectively), and the risk increased continuously with increasing UAE. The initial UAE thus emerged as an efficient and independent predictor of cardiovascular disease in middle-aged hypertensive and normotensive men. UAE appeared to be a stronger predictor than blood pressure and serum cholesterol.