The Hypertension Detection and Follow-up Program followed up 10,940 persons for 5 years in a community-based, randomized, controlled trial of treatment for hypertension. Participants were randomized to one of two treatment groups, stepped care and referred care. The primary end point of the study was all-cause mortality, with morbid events involving the heart, brain, and kidney as secondary end points. Loss of renal function, ascertained by a change in serum creatinine, was among these secondary events. Baseline serum creatinine concentration had a significant prognostic value for 8-year mortality. For persons with a serum creatinine concentration greater than or equal to 1.7 mg/dl, 8-year mortality was more than three times that of all other participants. The estimated 5-year incidence of substantial decline in renal function was 21.7/1,000 in the stepped-care group and 24.6/1,000 in the referred-care group. Among persons with a baseline serum creatinine level between 1.5 and 1.7 mg/dl, the 5-year incidence of decline was 113.3/1,000 (stepped care) and 226.6/1,000 (referred care) (p less than 0.01). The incidence of decline in renal function was greater in men, blacks, and older adults, as well as in those with higher entry diastolic blood pressure. Among persons with a baseline serum creatinine level greater than or equal to 1.7 mg/dl, serum creatinine concentration declined by 25% or more in 28.6% of stepped-care and 25.2% of referred-care participants. Although the incidence of clinically significant hypercreatininemia in a hypertensive population is low, an elevated serum creatinine concentration is a very potent independent risk factor for mortality. The slightly lower rate of development of hypercreatininemia and the higher rate of improvement in stepped-care compared with referred-care participants is consistent with the belief that aggressive treatment of hypertension may reduce renal damage and the associated increased risk of death.