The incidence of true scaphoid fracture and the value of radionuclide scan was evaluated in a 1-year retrospective review of 54 emergency department patients with a preliminary diagnosis of clinical scaphoid fracture (snuff box tenderness and negative initial X-ray study). All emergency department charts of patients with a discharge diagnosis of clinical scaphoid fracture were reviewed to determine the number ultimately found to have a true scaphoid fracture and to ascertain how the diagnosis was actually achieved. All patients were managed by immobilization in a thumb spica cast for a period of 10-14 d, with repeat clinical assessment and radiographs at that time. The diagnosis of true scaphoid fracture was confirmed radiographically at the time of reassessment or, in the case of persistent tenderness and negative repeat radiographs, by the use of a technetium bone scan. In our study group, only 8 of 54 patients (14.8%) were found to have a final diagnosis of true scaphoid fracture. In 75% (6/8) of these patients, the diagnosis was made by technetium bone scans, as the radiographs remained negative at the time of reassessment. In addition, a bone scan revealed the presence of two previously unsuspected fractures (one distal radius fracture and one triquetral fracture) in two patients with continued clinical tenderness. This study confirms the previously reported low incidence of scaphoid fractures in patients with a diagnosis of clinical scaphoid fracture. It also demonstrates the importance of technetium bone scan as a diagnostic tool of increased sensitivity, as compared to traditional radiographs, in detecting the presence of scaphoid fractures and in detecting other clinically unsuspected fractures in this group of patients.