Lethal carcinomas are still found inadvertently in patients under surveillance; some may not be preceded by conventional dysplasia. However, there is a survival advantage for cancers detected endoscopically rather than symptomatically, and, therefore, by preventing them by colectomy when dysplasia first becomes apparent. It may, therefore, be unnecessary to grade dysplasia when found, for if unequivocally present, then immediate consideration of colectomy is appropriate. It is unreasonable to expect colonoscopists to rebiopsy what might be a minute patch of dysplasia that has no distinguishing features endoscopically. Aneuploidy deserves consideration as a potential marker of patients at particular risk of developing dysplasia, who might undergo more frequent colonoscopy and biopsies than those without the presence of aneuploidy. There is still considerable interobserver variability in the grading of dysplasia by pathologists; part of this may be because grading occurs around a mean, the width of the tails of this distribution curve determining interobserver variation.