This method was designed to quantify root resorption on human premolar root surfaces induced by orthodontic forces by volume. Light (25 g) or heavy (225 g) orthodontic forces were applied to 20 first maxillary premolars in 10 human subjects. The contralateral teeth of the subjects served as controls. All teeth were extracted after 28 d of experimentation and prepared for imaging. A pair of stereo scanning electron microscopy (SEM) images (± 3°) of resorption craters was captured and imported into an image analysis software package. The images were aligned and grayscale depth maps of the craters were generated. Correction for errors due to residual tilt and curvature of the cementum surface using shading correction was performed. Thresholding was used to obtain a measure of both the cementum surface height and the average depth of the crater. The depth of the crater was the difference in these values. Crater volumes were obtained by multiplication of the average of this difference by area of the crater. Calibration of this volumetric measurement against standardized calculated known volumes on metallic rods showed good accuracy and reproducibility. In the experimental teeth, heavy forces caused threefold more resorption than light forces (P < 0.01). There was also more root resorption evident in the experimental teeth compared with the control teeth in both the light and heavy force groups.