Summary In a population-based mammography screening, 129 731 examinations were carried out among 36 000 women aged in the city of Turku, Finland, in the period 1987-94. Women older than 50 were screened at 2-year intervals, and those younger than 50 at either 1-year or 3-year intervals, depending on their year of birth. Screen-detected breast cancers numbered 385 and, during the same time period, 154 women were diagnosed with breast cancer outside screening in the same age group in the same city, and 100 interval cancers were detected. Two hundred and fifty (67%) of the screen-detected cancers were of post-surgical stage compared with 45 (45%) of the interval cancers and 52 (34%) of the cancers found outside screening (P<0.0001). However, among women aged 40-49 the frequency of stage cancers did not differ significantly among screen-detected cancers, interval cancers and cancers found outside screening (50%, 42% and 44% respectively; P=0.73). Invasive interval cancers were more frequent among women aged 40-49 if screening was done at either 1-year (27%) or 3-year intervals (39%) than in older women screened at 2-year intervals (18%; P=0.08 and P=0.0009 respectively). Even if adjusted for the primary tumour size, screen-detected cancers had smaller S-phase fractions than interval cancers or control cancers (P=0.01), but no difference in the S-phase fraction size was found between cancers of women younger than 50 and those older than this (P=0.13). We conclude that more interval cancers were found among women younger than 50 than among those older than 50 and that this could not be explained by the rate of cancer cell proliferation.Keywords: breast cancer; mammography; flow cytometry Mammography is currently used for population-based mass screening of breast cancer in several countries because it can detect asymptomatic breast cancer at an early stage when dissemination of cancer is still unlikely. Small asymptomatic cancers can be detected by mammography not only among women older than 50 years, but also among younger women aged 40-49 years (Peeters et al, 1989;Ikeda et al, 1992;Moss et al, 1993;Burhenne et al, 1994). In a meta-analysis screening mammography reduced breast cancer mortality by 26% (95% CI 17-34%) in women aged 50-74 but did not significantly reduce breast cancer mortality in women aged 40-49 (Kerlikowske et al, 1995). However, analysis of randomized trials on mammography screening concluded that, if the Canadian National Breast Screening Study was excluded from the analysis, a statistically significant benefit of 23% was present, in favour of the screened women . It is debatable whether or not population-based mass screening should be carried out among women younger than 50 (Fletcher et al, 1993;Kaluzny et al, 1994).Although a shorter screening interval than 2 years has been recommended and used by some (Morrison et al, 1988;Tabar et al, 1995), a meta-analysis of studies in which mammography screening at various intervals was compared with no screening failed to show that screening at 12-month in...