The bleeding time is used by many nephrologists to predict risk of hemorrhage before percutaneous kidney biopsy. Developed in 1910, the bleeding time is a nonspecific test that may be prolonged in multiple disease states. When accompanied by a platelet count, hematocrit, and a thorough investigation of family or personal history of bleeding, the bleeding time is the best predictor of hemorrhagic risk in patients with kidney disease. Because there is a small but significant risk of bleeding with percutaneous kidney biopsy, a prolonged bleeding time should be treated with 1-deamino-8-D-arginine vasopressin, cryoprecipitate, estrogens, or dialysis as indicated before biopsy. Treating all patients with 1-deamino-8-D-arginine vasopressin without checking bleeding times may be cost-ineffective when compared with treating only those patients with prolonged bleeding times.