This paper reports the sensitivity and specificity of self-reported ABO blood type. Subjects were 283 outpatientswho participated in Helicobacter pylori eradication program at Aichi Cancer Center Hospital. Excluding sevenpatients, an ABO blood test was successfully conducted for 276 participants (133 males and 143 females). Only threeparticipants answered their blood type differently from the results of the blood test; one A-type male answered to bewith AB type, one B-type male with AB type, and one AB-type male with A type. Nine participants stated that theydid not know their ABO blood type. The sensitivity was 98.7% (95% confidence interval, 92.8-100.0%) for 75 Otypeindividuals, 96.1% (90.3-98.9%) for 102 A-type individuals, 92.9% (84.1-97.6%) for 70 B-type individuals, and93.1% (77.2-99.2%) for 29 AB-type individuals, when the nine individuals were included. The specificity was morethan 99% for any blood type. These findings revealed that self-reported ABO blood type is highly valid forepidemiologic studies.