The Missouri Breast Cancer Detection Demonstration Project (BCDDP) at the Cancer Research Center in Columbia, Missouri, detected 136 malignant lesions among 10,187 asymptomatic participants between 1974 and 1980. Mammography was the sole detection modality in 76 or 56% of cases (Group 1). Mammography combined with physical examination revealed 41 (30%) additional cases (Group 2). Physical examination alone detected 19 (14%) cases (Group 3). Follow-up was complete in all cases. There were 26 (19%) noninvasive and 110 (81%) invasive lesions. Positive axillary nodes were found in 34 (25%) cases, and 102 (75%) patients had negative nodes. The overall 5-year survival was 84.5%. After a minimum follow-up of 5 years and a median follow-up of 8.5 years, 108 patients are alive, 5 of whom have recurrences. Among 28 deceased patients, 9 died of causes unrelated to breast cancer. The 5-year survival of Group 1 was 95% (72/76). Among this group of patients, 84% (64/76) had negative lymph nodes. At 5 to 10 years follow-up, there have been no breast cancer-related deaths among this group, and only one patient has recurrent disease 91 months after mastectomy. Furthermore, 85% (22/26) of all noninvasive lesions were found in this group. Of the remaining 12 patients diagnosed by mammography alone but who had positive nodes, the 5-year survival was 75% (9/12). By contrast, patients diagnosed by mammography plus physical examination (Group 2) experienced a survival of 71% (29/41) at 5 years. Sixty-three percent of patients diagnosed by physical examination had negative lymph nodes. Their 5-year survival was 84%, which contrasts with a 50% 5-year survival for patients with positive nodes. In Group 3, 79% (15/19) had negative nodes. The 5-year survival rate of this group was 74% (14/19), and the 5-year survival rate of the node-negative patients with invasive disease in this group was 29% and 21% lower than that of the node-negative patients with invasive disease of Groups 1 and 2, respectively (P less than 0.01). This study suggests that improved survival in breast cancer screening program is in large measure dependent on the ability to detect lesions before they become palpable. The authors conclude that mammography is a determining factor in early diagnosis, and at 5 to 10 years contributes significantly to improved 5- and 10-year survival rates.