An epidemiological evaluation of the protective effect for dying of stomach cancer by screening programme for stomach cancer was conducted with applying a method of case-control study. And also in order to carry out an efficient screening programme, the age groups who should be intensively recommended to receive screening and an optimal screening time interval since the last test were analyzed. 527 cases of men and 273 of women, dying of stomach cancer in the years 1981-1989 in the Awa region of Chiba prefecture, were identified from Chiba Cancer Registry. For each case, 3 controls were drawn at random from Awa living residents (about 162000 inhabitants at 1989 national census), with being matched strictly according to the district of residence, sex and born within 3 years of birth-year. For both cases and controls, the information about the screening history until the date of diagnosis of the case in each matched set was collected respectively from comparison with the screening certification. The results showed a relative risk of 0.417 (99% CI 0.284-0.612) in ever screened men compared with never screened and 0.480 (99% CI 0.280-0.823) in women. The significant reduction in risk was intensively observed on age groups 40-74 years among men and 50-69 years among women and the protective effect continued at most in the following three years since last screening. For an efficiency of screening programme, these age groups should be intensively recommended to receive screening and it is allowable that an optimal screening time interval since last negative test is at most 3 years for general attendance.