Investigators can infer how much reduction in volume has occurred since brain volume was at its peak, by combining measures of brain volume with measures of intracranial volume (ICV). Several methodologies have been proposed to asses the ICV. However, we have not seen a gold-standard study evaluating the results of the methodologies for the assessment of ICV. In the present study, the actual intracranial volume of 20 dry skulls was measured using the water-filling method, using this as a gold standard. Anthropometry, cephalometry, point-counting, and planimetry techniques were applied to the same skulls to estimate the ICV. Anthropometric and cephalometric measurements were carried out directly on skulls and roentgenograms, respectively. Consecutive computed tomography sections at a thickness of 10 mm were used to estimate the ICV of the skulls by means of the point-counting and planimetry methods. The mean (+/- SD) of the actual ICV measured by the water-filling method was 1,262.0 +/- 160.4 cm(3) (1,389.5 +/- 96.5 cm(3) for males and 1,134.5 +/- 94.3 cm 3 for females, respectively). Our results showed that the estimated values obtained by all four methods differed from the actual volumes of the skulls (P < 0.05). The data obtained by anthropometry resulted in overestimation. However, cephalometry, point-counting, and planimetry methods produced underestimation. After calibration, there were no significant differences between the actual volumes and the results of the four methods (P > 0.05). While the anthropometric method is easy and quick to apply, its result may deviate from the actual values. The optimized stereological techniques of point-counting and planimetry methods may provide unbiased ICV results since they take the third dimension of the structures into account.