Plagiocephaly is a descriptive term that connotes an asymmetrically oblique or twisted head. Such cranial dysmorphology has a number of etiologies, the most common of which are unicoronal synostosis, unilambdoid synostosis, and plagiocephaly without synostosis. Use of the term plagiocephaly in the literature is often ambiguous in that at times it is used inclusively for all etiologies while at other times it is used exclusively as a synonym for unicoronal synostosis. Although differentiation by physical examination among unicoronal synostosis, unilambdoid synostosis, and plagiocephaly without synostosis usually is possible for an experienced observer, inexperienced observers often have difficulty making an anatomically accurate diagnosis even with the assistance of conventional skull radiographs. High-resolution CT scans, including three-dimensional osseous surface re-formations, have become a standard element in the evaluation of craniofacial anomalies in many centers. We hypothesized that the three major etiologies of plagiocephaly could be unambiguously differentiated by means of endocranial three-dimensional CT osseous surface re-formations. Archival pretreatment CT data on 15 unicoronal synostosis, 4 unilambdoid synostosis, and 15 plagiocephaly without synostosis patients were reviewed to define, qualitatively and quantitatively, the characteristics of the endocranial base morphologies for each group; in addition to visual dysmorphology specific to each group, there was a statistically significant difference in the angle of deviation from the midlines of the anterior and posterior cranial fossae among unicoronal synostosis, unilambdoid synostosis, and plagiocephaly without synostosis. Four radiologists experienced in reading images of craniofacial anomalies were oriented to the group characteristics and then instructed to perform differential diagnosis for each of the 34 patients using only the endocranial three-dimensional CT images. The raters were blind to all other clinical and diagnostic information. The raters correctly diagnosed unicoronal synostosis. Errors were made in differentiation of unilambdoid synostosis and plagiocephaly without synostosis. These errors resulted from the raters' reliance on image inspection rather than quantitation of anteroposterior fossae midline angulation. Such quantitation unambiguously differentiated between unilambdoid synostosis and plagiocephaly without synostosis in the "error" cases. The endocranial base dysmorphology of patients with plagiocephaly is etiology-specific for unicoronal synostosis, unilambdoid synostosis, and plagiocephaly without synostosis. Three-dimensional CT endocranial base images can assist differential diagnosis of plagiocephaly.