To assess the reliability and validity of osteophony (patellar-pubic percussion [PPP] test) as a physical diagnostic sign in the evaluation of hip trauma, a prospective study was undertaken of 41 consecutive patients presenting to the emergency department with a history of hip trauma necessitating radiographic examination. Fifteen of 19 (78.9%) patients who presented with a history of hip trauma and a fracture on radiograph were found to have had an abnormal PPP sign by at least one of two raters (P < .0001). Only 1 of 22 (4.6%) patients without evidence of fracture (eg, contusion) had an abnormal PPP sign. This patient had diffuse Paget's disease. Nine of 10 (90%) patients who had trochanteric fractures had an abnormal PPP sign (P < .02). Overall reliability of the PPP sign based on two observers was 90.2% (P < .0001). In those patients with radiographic evidence of fracture, interrater reliability was 84.2% (P < .0001). For patients in whom physicians agreed on the PPP sign, the PPP test resulted in a 0% false-positive error and a 25% false-negative error. For patients in whom either physician noted an abnormal PPP sign, the PPP test resulted in a 4.6% false-positive error and a 21.1% false-negative error. The presence of an abnormal PPP sign in the evaluation of hip trauma is associated with evidence of fracture or other bony abnormality on radiograph.