The diagnostic accuracy of telepathologic analysis has not been compared to that of conventional light microscopic review on a difficult case consultation service. The anatomic pathology consultation files of the University of Iowa were retrospectively examined, and 105 difficult cases from a variety of organs were chosen for real-time telepathologic and light microscopic review. The telepathologic and light microscopic crude agreement of five pathologists were compared, with use of the original consultation diagnosis as the "gold standard." Cases were scored as correct, partially correct, or incorrect. After making a video diagnosis, the pathologists reported whether they wanted to review the case with use of a light microscope. The pathologists performed significantly better with the light microscope, even after excluding cases in areas of inexpertise (P = .005). The mean percentage of cases that the pathologists wanted to review with the light microscope was 64%, and the major reason for review was diagnostic uncertainty. Cases incorrectly diagnosed with use of the video monitor were almost always requested for review. We conclude that, on a difficult case consultation service, pathologists perform significantly better with use of light microscopic than with telepathologic analysis; rarely make an incorrect diagnosis and do not request that case for light microscopic review; and exhibit high telepathologic diagnostic accuracy in areas of expertise.