We reviewed the evidence relating to breast screening using mammography in younger women, primarily those aged 40 to 49. The best evidence comes from the 6 published randomized trials. We performed a metaanalysis, assessing the breast cancer death rates for each year of follow-up in each trial, and producing summary estimates of the cumulative and year-specific mortality rates. At 7 to 10 years, death rates in screened and unscreened groups are similar. This lack of benefit contrasts with the average 30% mortality reduction in women over age 50 in the same trials. This evidence provides no basis for the promotion of mammographic screening in women under age 50 in the general population. If mammography is to be offered to asymptomatic women of that age who request it, appropriate information including the fact that there has been no demonstrated ultimate benefit should be given. The difference in effectiveness may relate both to the greater difficulty in distinguishing normal from abnormal tissue, and to greater growth speed and different biological characteristics of tumors, in younger women. We also reviewed recommendations from major groups: in general multidisciplinary bodies which rely on structured evaluations do not recommend screening in younger women, while many other specialty associations and cancer societies do. Overall, the trials of screening of younger women show a small increase in mortality in those screened, and possible explanations of this are discussed. [corrected] Remaining questions concern the precise age at which effectiveness changes, the situation in high-risk women, and the impact of improved or more frequent screening.