To effectively study risk factors, a uniform definition of chronic rejection must be agreed upon. The definition must describe a typical clinical course--e.g., gradual deterioration in graft function starting at least 3 months post-transplant and lasting at least 6 months--in combination with a characteristic biopsy. Numerous kidney transplant studies indicate that an acute rejection episode is an important predictor of 1-year graft survival and half-life (the time it takes for half the grafts that survive the 1st year to subsequently fail). However, in each of these studies there are multiple potential causes of graft loss--acute rejection, chronic rejection, patient death, and recurrent disease. We have tried to eliminate this problem by studying patients with biopsy- or nephrectomy-proven chronic rejection. In a univariate analysis we found that an acute rejection episode significantly increased the risk of developing biopsy-proven chronic rejection (p < 0.001). Risk was higher in patients with more than one acute rejection episode and in patients whose rejection episode occurred more than 60 days after transplant. In a separate multivariate analysis, we studied the impact of the following variables on the incidence of biopsy-proven chronic rejection: transplant number (first or retransplant), age at transplant (over 18, 18 to 50 years old, under 50), gender, HLA-ABDR match, peak and transplant panel reactive antibody, number of acute rejection episodes, infection, donor age, and cyclosporine (CsA) dose at 1 year (less than 5 mg/kg versus more than 5 mg/kg).(ABSTRACT TRUNCATED AT 250 WORDS)