Although often recommended as an important component in screening for breast cancer, physical examination of the breasts (PE) by medical professionals has not been well evaluated. The Canadian National Breast Screening Study (NBSS) permits estimation of sensitivity, specificity, and positive predictive value (PPV) of PE alone as performed by screen-examiners because 50% of the 89,835 NBSS participants did not receive mammography. There were 19,965 women aged from 50 to 59 years who were eligible to receive four or five annual PEs from 77 nurse-examiners, in 12 screen centers outside Quebec province and 58 physician-examiners in three screen centers in Quebec. The gold standard was histologically proven breast cancer. When a test was positive the participant was referred to the study surgeon for review. For screens one to five, sensitivity was 83, 71, 57, 83, and 77; specificity was 88, 94, 96, 96, and 96; and PPV was 3, 3, 4, 3, and 4, respectively. For 25,620 women aged 40 to 49 years who were eligible to receive only one PE, sensitivity was 71, specificity 84, and PPV 1.5. Using a binomial regression model, X2 for heterogeneity suggested there was no difference between nurse and physician examiners (P = 0.6879). Similar estimates made for the surgeons who performed 8914 reviews showed that sensitivity and PPV were higher than for the screen examiners and specificity was lower. These results support the conclusion that physical examination of the breasts by trained nurses is a useful component in screening for breast cancer.