The purpose of this study was to evaluate the quantitative accuracy of three-dimensional (3D) rendered images acquired with multi-detector row computed tomography (MDCT) by means of distance measurements of a dry human skull for various slice thicknesses and acquisition modes. A radiologist directly measured the distance of line items on the skull surface to establish reference "gold standards." The skull specimen was scanned with a MDCT with various scanning parameters (slice thicknesses and acquisition modes). An observer measured the corresponding distances of the same items on 3D rendered images. The quantitative accuracy of distance measurements was statistically evaluated. There were no significant statistical differences (P value <.05) in accuracy of distance measurements among the scan modes. However, the results showed that acquisition slice thickness was the influential factor in determining the accuracy of the 3D rendered MDCT images. The quantitative analysis of distance measurement may be a useful tool evaluating the accuracy and defining optimal parameters of 3D rendered images.