This study is aiming to determine the validity and observer variability of clinical examination and mammography as screening tests for breast cancer. Women over the age of forty are given two independent clinical examinations of the breasts, and mammograms are taken and read independently by two radiologists. This paper presents the results of screening the first 1215 women to be enrolled in the study. At their first screening attendance, 231 women (19%) were referred for surgical opinion, 119 (9-8%) underwent biopsy, and cancer was diagnosed in 17 (1-4%). 2 further cancers were diagnosed in the ensuing six months among women who had been negative on initial screening, representing a false-negative rate of 2 out of 19 (11%). Clinical examination resulted in 189 referrals (15-6%), 90 biopsies (7-4%), and detected 11 cancers; corresponding figures for mammography were 76 referrals (6-3%), 55 biopsies ((4-5%), and 14 cancers. Observer variability was greater for clinical examination than for mammography. These early results suggest that as a screening test mammography compares favourably with clinical examination, but both tests are necessary if many false negatives are to be avoided.