All inhabitants of the city of Göteborg who in 1982 were between 60 and 64 years of age (27,700) were randomly divided into a test and a control group. The 13,759 subjects in the test group were invited to perform Hemoccult II (Smith Kline Diagnostic, Sunnyvale, CA) fecal occult blood testing over 3 days and to repeat the testing after 16 to 22 months. At the first screening 9,040 (66%) completed the test, and 7,770 (58%) completed the test at the second screening. In the first screening the test group was divided into two subgroups in which the tests were rehydrated and unhydrated before development. All tests were rehydrated in the second screening; 1.9% and 5.8% of the tests were positive in the unhydrated and rehydrated subgroups, respectively. The number of diagnosed neoplasms in the first screening was significantly larger (P less than 0.01) in the rehydrated group compared to the unhydrated group, 50 and 24 neoplasms, respectively. Sixteen of 61 carcinomas in the test group were found in the interval between the two screenings, 19 of the carcinomas at the second screening, and ten among the nonresponders. Rehydration of the Hemoccult II test is a necessity. Significantly more carcinomas (61) were found in the test group compared to the control group (20). There was a trend toward favorable tumor staging in the test group compared to the control group.