Previous publications have reported an unusually high colon cancer mortality rate for several Kentucky counties. We investigated these high rates by examining incidence of colorectal cancer in one county with a high mortality. The objective was to determine whether the incidence of colorectal cancer was as high as mortality rates indicated and, if so, to look for possible etiologic factors for the high rates. We found the incidence of colon cancer to be significantly higher in Campbell County than expected. While we expected 162 cases of colon cancer, we actually observed 192 (P less than .01). The number of rectal cancers was no higher than expected (52 expected and 62 observed), in agreement with previously reported mortality figures. A geographic plot of cases by home residence showed a significantly higher rate of colon cancer for urban county regions than for rural regions. In fact, the population of rural Campbell County had a colon cancer rate significantly lower than either the county rate or the national rate. Several factors were analyzed to explain these rate differences. The only consistently associated factor was source of residential drinking water.