Despite the abundance of literature on nonoperative management (NOM) of blunt trauma to the liver and spleen there is limited information on NOM of blunt renal injuries. In an effort to evaluate the role of NOM 37 consecutive unselected patients with renal injuries (grade 1, four; grade 2, 12; grade 3, 11; grade 4, six; and grade 5, four) were followed prospectively over 30 months (Match 1999 to September 2001). Patients without peritonitis or hemodynamic instability were managed nonoperatively regardless of the appearance of the kidney on CT scan. Six (16%) patients were operated on immediately but only two (5.4%) for the kidney (grades 3 and 5 respectively). Of the remaining 31 patients 26 (84%) were managed successfully without an operation (grade 1 or 2, 12; grades 3-5, 14). Five patients were taken to the operating room after a period of observation (3, 3.5, 9, 36, and 44 hours respectively) but only three for the kidney (grades 4 and 5). The overall failure rate was 16 per cent (5 of 31); the rate of failure specifically related to the renal injury was 9.6 per cent (three of 31). Compared with the patients with successful NOM the five patients with failed NOM were more severely injured (Injury Severity Score > or = 15 in 80% vs 27%, P = 0.04), required in the first 6 hours more fluids (4.17 +/- 1.72 vs 1.87 +/- 1.4 liters, P = 0.003) and blood transfusions (2.40 +/- 2 vs 0.42 +/- 1.17 units, P = 0.005), and more frequently had a positive trauma ultrasound (80% vs 11.5%, P = 0.005). We conclude that NOM is the prevailing method of treatment after blunt renal trauma. It is successful in the majority of patients without peritonitis or hemodynamic instability and should be considered regardless of the severity of renal injury. Predictors of failure may exist on the basis of injury severity, fluid and blood requirements, and abdominal ultrasonographic findings and need validation by a larger sample size.