Hospitalization for observation is the current standard of practice for patients who have sustained blunt abdominal trauma and who do not require emergent operation, despite having undergone diagnostic studies that exclude the presence of an intra-abdominal injury. The reasons for this practice are multifactorial and include the perceived false-negative rate of all standard diagnostic tests, the belief that hospitalization will allow for the prompt diagnosis of occult injuries, and medicolegal considerations about the risk of early discharge. The focus of this study was to determine whether hospitalization for observation is necessary after a negative diagnostic evaluation after blunt abdominal trauma, to determine the negative predictive value of abdominal computed tomographic (CT) scanning in a prospective series of patients, and to identify which patients can be safely released from the emergency department without observation or hospitalization after blunt abdominal trauma. In a prospective, multi-institutional study over 22 months at four Level I trauma centers, all patients with blunt abdominal trauma suspected by either physical examination or mechanism of injury were evaluated using the following protocol: physical examination in the emergency department, followed by abdominal CT scanning, followed by hospitalization for observation. The standardized physical examination was repeated between 4 and 8 hours. Outcomes were measured at 20 hours and at discharge and included clinical deterioration, the need for celiotomy, and mortality. Other data collected included demographics, mechanism of injury, and findings on physical examination and abdominal CT scanning. Three thousand eight hundred twenty-two consecutive patients with suspected abdominal trauma presented to the four trauma centers. Two thousand seven hundred seventy-four of these met study eligibility criteria and were prospectively enrolled. Of these, 2299 fulfilled the entire study protocol. CT scan was negative in 1,809 patients, positive for organ injury or abdominal fluid in 389 patients, and nondiagnostic in 78 patients. Abdominal tenderness or bruising was present in 1,380 patients (61%), but only 22% had a positive CT scan. Nineteen percent of patients with a positive CT scan had no tenderness. Computed tomography detected 22 of the 25 blunt intestinal injuries in this series. Free intraperitoneal fluid without solid visceral injury was present in 90 patients, and but only 7 patients had intestinal injuries. There were nine celiotomies in patients whose CT scan was initially interpreted as negative: six were therapeutic (intestine in three, bladder in one, kidney in one, and diaphragm in one), two were nontherapeutic, and one was negative. The negative predictive power of an abdominal CT scan based on the preliminary reading and as defined by the subsequent need for a celiotomy in the population fully satisfying the protocol was 99.63% (lower 95 and 99% confidence bounds of 99.31 and 99.16%, respectively). These data indicate that abdominal tenderness is not predictive of an abdominal injury and that patients with a negative CT scan after suspected blunt abdominal trauma do not benefit from hospital admission and prolonged observation.