Background: The potential role of digestive endoscopy as a mode for transmissi on of hepatitis C virus (HCV) is controversial. Objective: To evaluate the role of digestive endoscopy in transmitting HCV by comparing the incidence of HCV inf ection in a cohort of patients undergoing endoscopy and in a cohort of blood don ors. Design: Prospective cohort study. Setting: 3 endoscopic units and 2 blood b anks in northwestern Italy. Patients: The potentially exposed cohort consisted o f 9188 out-patients consecutively recruited from 3 endoscopic units. Of 9008 pa tients negative for antibody to HCV (anti-HCV), 8260 (92%) were retested for a nti-HCV 6 months after endoscopy. The unexposed cohort consisted of 51 230 heal thy, anti-HCV-negative persons who donated blood at 2 blood banks in the same area and during the same time period; 38 280 of them (75%) were tested again fo r anti-HCV 6 to 48 months after the first blood donation (95 317 person years o f observation). Measurements: Differences in the anti-HCV seroconversion rate b etween the exposed cohort (patients undergoing endoscopy) and the unexposed coho rt (blood donors). Seroconversion was evaluated by a third-generation enzyme im munoassay for anti-HCV; persons positive for anti-HCV were tested for HCV RNA by polymerase chain reaction. Results: All 8260 persons undergoing endoscopy rem ained negative for anti-HCV 6 months after the procedure (risk per 1000 persons , 0 [95%CI, 0 to 0.465]); in particular, none of the 912 patients who underwent endoscopy with the same instrument previously used on HCV carriers showed anti -HCV seroconversion (risk per 1000 persons, 0 [CI, 0 to 4.195]). Four blood don ors became positive for anti-HCV and HCV RNA (mean follow-up, 2.49 years; 0.04 2 case per 1000 person-years [CI, 0.011 to 0.107 case per 1000 personyears]); e ach had undergone minor surgery before the second test. Limitations: In the endo scopy cohort, 8.3%of patients were lost to follow-up. Conclusions: These findi ngs support the hypothesis that properly performed digestive endoscopy is not a major risk factor for the transmission of HCV. Background: The potential role of digestive endoscopy as a mode for transmissi on of hepatitis C virus (HCV) is controversial. Objective: To evaluate the role of digestive endoscopy in transmitting HCV by comparing the incidence of HCV infraction in a cohort of patients undergoing Endoscopy and in a cohort of blood don ors. Design: Prospective cohort study. Setting: 3 endoscopic units and 2 blood b anks in northwestern Italy. Patients: The potentially exposed cohort consisted of 9188 out-patients consecutively recruited from 3 endoscopic units. Of 9008 pa tients negative for antibody to HCV (anti-HCV), 8260 (92%) were retested for a nti-HCV 6 months after endoscopy. The unexposed cohort consisted of 51 230 heal thy, anti-HCV-negative persons who donated blood at 2 blood banks in the same area and during the same time period; 38 280 of them (75%) were tested again fo anti-HCV 6 to 48 months after the first blood donation (95 317 person years of observation). Measurements : Difference s in the anti-HCV seroconversion rate b etween the exposed cohort (patients undergoing endoscopy) and the unexposed coho rt (blood donors). Seroconversion was evaluated by a third-generation enzyme im munoassay for anti-HCV; persons positive for anti-HCV were tested for HCV RNA by polymerase chain reaction. Results: All 8260 persons undergoing endoscopy remined negative for anti-HCV for 6 months after the procedure (risk per 1000 persons, 0 [95% CI, 0 to 0.465]); none of the 912 patients who underwent endoscopy with the same instrument previously used on HCV carriers showed anti-HCV seroconversion (risk per 1000 persons, 0 [CI, 0 to 4.195]). Four blood donors became positive for anti-HCV and HCV RNA (mean follow-up, 2.49 years; 0.04 2 case per 1000 person-years [CI, 0.011 to 0.107 case per 1000 personyears]); ech had undergone minor surgery before the second test. Limitations: In the endo scopy cohort, 8.3% of patients were lost to follow-up. Conclusions: These findi ngs support the hypothesis that properly performed digestive endoscopy is not a major risk factor for the transmission of HCV.