To determine the parameters associated with significant bleeding and to examine the value of performing a renal biopsy, we studied 83 consecutive patients, including 24 renal allograft recipients, who had undergone percutaneous renal biopsy. The patients were stratified into four groups according to the percentage of decline in their hematocrit (Hct) at 24 hours postbiopsy, as follows: 10% or greater (n = 21; 25%) and less than 10% decline (n = 62; 75%). The latter group was further subgrouped into 5% to 10% (n = 22) and less than 5% decline (n = 40). There was a significant decline in Hct postbiopsy, with a linear correlation between the decrease in Hct at 6 and 24 hours (R2 = 0.47; P < 0.0001), suggesting that the former was a predictor of the latter. There was a linear correlation between the number of passes and number of cores obtained for the first four passes, but an inverse correlation when five passes or greater were required. Interestingly, there was no correlation between bleeding (>10% decline in Hct) and the number of passes or cores obtained. Gross hematuria and blood transfusion requirement were each encountered in three patients (3.6%). Importantly, the prebiopsy clinical diagnosis was altered in 18 of 59 native kidney biopsies (33%) and 10 of 24 transplant biopsies (41%). We conclude that percutaneous renal biopsy using an automated spring-loaded gun device coupled with ultrasound guidance is a safe technique and provides essential clinical information. Importantly, patients with a stable Hct at 6 hours were at low risk for bleeding at 24 hours while hospitalized. It remains to be determined if these findings could be extrapolated to early discharge from hospital.