We investigated the predictive power of morphological features in 224 autistic patients and 224 matched-pairs controls. To assess the relationship between the morphological features and autism, we used the receiver operator curves (ROC). In addition, we used recursive partitioning (RP) to determine a specific pattern of abnormalities that is characteristic for the difference between autistic children and typically developing controls. The present findings showed that morphological features are significantly increased in patients with autism. Using ROC and RP, some of the morphological measures also led to strong predictive accuracy. Facial asymmetry, multiple hair whorls and prominent forehead significantly differentiated patients with autism from controls. Future research on multivariable risk prediction models may benefit from the use of morphological features.