Our goal was to investigate the effect of displayed image magnification on perception of the size of hepatic lesions on abdominal computed tomography (CT) scans. Institutional review board approval and informed observer consent were obtained. Three experienced radiologists reviewed 90 CT image pairs in one session. Each image pair demonstrated a solitary, well-defined hypodense hepatic lesion measuring greater than 1 cm obtained at two points in time. The image pairs were presented three times in random order, once with the left image magnified, once with the right image magnified, and once with neither image magnified. The radiologists were asked to determine on which image the lesion was smaller or if there was no difference. The responses were analyzed statistically. The proportion of correct responses increased significantly as the difference in lesion size increased (p < 0.001). The percent of correct responses was higher when neither CT image was magnified. Magnification of one image decreased the accuracy of the readers' performance, especially at smaller differences, both of which were statistically significant (p < 0.001). Thus, accuracy of detecting lesion size differences was degraded when the images were presented at differing magnification. This should be kept in mind when evaluating serial CT scans for growth or regression of tumors and other lesions.