The results of the Swedish two-county trial of mammographic screening for breast cancer are presented, updated to December 31, 1990, which is an average of 10.8 years follow-up per person. The main result of the trial in terms of breast cancer mortality remained the same: compared with the control group, the group invited to screening had a relative breast cancer mortality of 0.70 (P = 0.0002) with 95% confidence interval (0.58, 0.85). Analysis of survival showed that relative to the control group, the cancers detected at prevalence screen, incidence screens, and in the interval between screens had a good prognosis, whereas cancers detected in those who had refused screening had a very poor prognosis. When adjusted for tumor size, lymph node status, and tumor grade (differentiation), the better survival associated with incidence screen and interval detection was largely accounted for, indicating that the benefit of incidence screening is largely achieved through the effect of screening on the three prognostic variables, notably size of the tumor. Results indicate that to achieve a substantial mortality reduction, 50% of screen-detected invasive cancers should be less than 15 mm in diameter, at least 30% of screen-detected grade 3 tumors should be less than 15 mm, and at least 70% of screen-detected tumors should not have lymph node metastases. The percentage of grade 3 tumors of a given size should be the same in screen-detected cancers as in clinically detected, and breast cancer prevalence at first screen should be at least three times the expected incidence rate in the absence of screening. This should be achieved without the recall rate for further examination exceeding 9%, and procedures including further imaging techniques and fine needle aspiration or core biopsy should be used before resorting to surgical biopsy. These aims can be achieved in specialist screening centers if particular attention is paid to resources for screening and diagnostic evaluation, specialist training of clinical and technologic screening staff, and ongoing monitoring of mammographic quality, recall rates, and the attributes of the tumors detected.