The objective of this study was to compare the sensitivity and specificity of 2- and 4-film x-ray series when interpreted by pediatric emergency medicine physicians in the diagnosis of skull fracture in children. A noninferiority crossover study was performed. The skull radiographs of the 50 most recent cases of skull fracture for which a 4-film radiography series was available and 50 controls matched for age were reviewed. Two modules, containing a random sequence of 2- and 4-film series of each child, were constructed to have all children evaluated twice (once with 2 films and once with 4 films). Pediatric emergency physicians evaluated both modules 2 to 4 weeks apart. The interpretation of the 4-film series by a pediatric radiologist served as the criterion standard. The sensitivity and specificity of the 2-film versus the 4-film skull x-ray series, in the identification of fracture, were compared. Thirteen pediatric emergency physicians participated in the study. For sensitivity, the mean difference between the 2- and 4-view series was higher than the noninferiority margin of 0.055 with an absolute mean difference of 0.060 (4-view minus 2-view series) and a 1-sided 95% higher confidence limit of 0.099. However for specificity, the mean difference was within the margin with an absolute mean difference of 0.011 and a 1-sided 95% higher confidence limit of 0.033. For children sustaining a head trauma, the 2-film skull radiography series is not as sensitive as the 4-film series in the detection of fracture, when interpreted by pediatric emergency physicians.