The ratio (R) of prevalence of screening-detected breast cancer in the first screening round (P) was compared with the expected incidence rate (I) for different age groups in several screening programs. Published data on the first screening round from three Swedish randomized trials and six counties with service screening were used. The women invited to take part in the screening were aged 40-74 years. Not only P and I but also R increased with increasing age. With the youngest age group as reference, the increase was statistically significant for both invasive cancer and invasive cancer and carcinoma in situ together. The studied ratio (R) can be thought of as a measure of efficiency in detecting breast cancer cases in mammography screening. The reasons for the increase are probably that the breast tissue of younger women is denser, which makes the cancer more difficult to detect by mammography, and that slow-growing cancers tend to appear more frequently in older women.