In 1973, 1 to 30 million asbestos-like fibers per liter of tap water were discovered in Duluth drinking water. Previous studies had linked mesothelioma, lung, and gastrointestinal cancers with occupational exposure to asbestos, so surveillance of cancer morbidity in Duluth was initiated to investigate effects from ingestion of asbestos in drinking water. Gastrointestinal and lung cancer incidence data for 1969-1974 were collected in the same manner as in the Minneapolis-St. Paul component of the Third National Cancer Survey; Duluth rates for 1969-1971 were compared with incidence rates for the cities of Minneapolis and St. Paul during the same time period; and Duluth rates for 1972-1974 were compared with Duluth rates for 1969-1971. Duluth females and both sexes combined had statistically significantly higher rates of pancreatic cancer than in Minneapolis and St. Paul in 1969-1971. These rates subsequently decreased in 1972-1974 for both sexes combined in Duluth. Duluth males and both sexes combined had similar excesses for gastrointestinal tract not specified in comparison with Minneapolis and St. Paul. Duluth and Minneapolis cancer incidence rates yielded less-exaggerated differences between the two study areas compared with mortality rates. Resources required for morbidity surveillance are described.