In response to a union request, a cancer screening program was conducted for the Pattern Makers' League of North America. Ten colon cancer cases were detected among the 1,465 white men screened with a flexible sigmoidoscope. The difficulties in obtaining appropriate “expected” numbers were that (1) prevalent detectable preclinical colon cancer is not equivalent to incident disease, and (3) the flexible sigmoidoscope yields results not directly comparable to those of the rigid sigmoidoscope used previously. The “expected” number of cancers was obtained by using an independent estimate of 5 years for the mean duration of the detectable preclinical phase. This implied that the expected number of colon cancer cases should be based on the age-specific incident rates among white men in the next-older 5-year age group and that the annual expected number should be multiplied by five. Therefore, the ten observed cases of colon malignancies represented an approximately threefold increase. For invasive cancer only, there was a slightly less than twofold cancer increase. Fifteen percent of the men had one or more colorectal polyps.