We prospectively compared clinical examination (CE) with plain films (PXR) and both tools with CT in patients sustaining blunt trauma. There were 1388 patients who had both PXR and CT of whom 168 (12.1%) were diagnosed with a fracture by CT. CE findings most predictive of fracture included age (OR, 1.025; CI, 1.011-1.039), hip pain (OR, 4.971; CI, 2.508-9.854), internal rotation of the leg (OR, 4.880; CI, 1.980-12.027), or tenderness to palpation over the sacrum (OR, 2.297; CI, 1.144-4.612), over the right or left hip (OR, 3.626; CI, 1.823-7.214), or diffusely throughout the pelvis (OR, 16.445; CI, 4.277-63.237). These factors were still predictive of pelvic fractures even in patients with a Glasgow Coma Scale score less than 13. There were 136 fractures identified by PXR all of which were identified by CE (sensitivity 100% [136 of 136], negative predictive value 100% [619 of 619]). There were six patients with negative clinical examinations and positive CTs (sensitivity 96.4% [162 of 168], negative predictive value 99.03% [613 of 619]), none of which were hemodynamically significant. The sensitivity for PXR compared with CT was 79.17 per cent (133 of 168) and the NPV was 97.2 per cent (1217 of 1252). CE is a reliable way to diagnose pelvic fractures and PXR is a poor screening tool for these injuries compared with CT. Because the majority of patients undergo CT after blunt trauma, routine screening radiographs should be eliminated.