We read the paper by Delvaux et al.  with interest. The study confirms the value of capsule endoscopy in the diagnosis of obscure gastrointestinal bleeding, and in this context also shows that capsule endoscopy is capable of identifying lesions that have not been detected in previous upper and lower endoscopies (10 nonneoplastic gastric and/or colonic lesions in 44 reported cases). The problem of missing lesions at endoscopy is particularly important in relation to tumors. Leaper et al.  reported a delayed diagnosis of colorectal cancer in 5.9 % of patients who underwent colonoscopy, potentially affecting their prognosis. These findings are very similar to those reported by Hosokawa et al. , who found a 6 % rate of delayed diagnosis, with a 3-year interval. In our own experience, including 59 capsule endoscopy examinations for obscure gastrointestinal bleeding carried out between January 2002 and December 2004, three polypoid lesions were found in the right colon that had not been detected at previous colonoscopies conducted elsewhere. Right hemicolectomy was subsequently carried out in all of the patients concerned, and the histology showed two adenocarcinomas and one tubulovillous adenoma with severe dysplasia. It is known that even when colonoscopy is conducted by a skilled endoscopist, there is still a significant miss rate for cancers. The miss rate at colonoscopy is higher with right-sided colon cancers (4 - 5 %) . The reasons for this high percentage include not only an inability to reach the ileocecal valve, but also inadequate bowel preparation, the presence of very small lesions, and the incorrect belief that one has reached the cecum. Before capsule endoscopy was introduced into clinical practice, the American Gastroenterological Association suggested that upper and lower endoscopies should be repeated in the presence of obscure gastrointestinal bleeding . In fact, however, when a colonoscopy is not clearly recognized as inadequate, significant delays in definitive diagnosis can occur, with a median duration of 10 months, as shown by Leaper et al. . Patients who are generally asymptomatic may often decline to undergo, or may delay, repeated endoscopies and/or other invasive procedures, with the risk of affecting the prognosis when a colon cancer has not been recognized. As Delvaux et al.  also show, capsule endoscopy in obscure gastrointestinal bleeding may lead to a significant reduction in health-care costs, since the subsequent endoscopy is aimed at classifying and taking a biopsy of the lesion. Cancers missed at colonoscopy could represent a major medicolegal issue. In this context, capsule endoscopy, as a noninvasive procedure, may be more acceptable to patients with obscure gastrointestinal bleeding and could play an important role in diagnosing missed lesions - particularly neoplastic ones. However, this will need to be confirmed by further large-scale studies.