To investigate the accuracy of video-microscopy of frozen sections, two pathologists reexamined 80 cases of archival material, on which frozen sections had previously been performed. Diagnoses based on the two trials (Observer 1 and 2) and diagnoses obtained in the original frozen section situation were compared with the final diagnoses based on paraffin embedded material. Observer 1 had two false negative diagnoses, but no false positive, whereas Observer 2 had one false positive, but no false negative diagnosis when compared with the final diagnoses. No false positive or false negative diagnoses were made in the original frozen sections situation and the number of inconclusive diagnoses were 5, as compared with 6 and 8 in the two trials based on video-microscopy. More experience with video-microscopy will probably achieve a quality of frozen section diagnoses similar to that of direct light-microscopy.