An ecologic study design was used to investigate the relationship between cancer risks and residence in communities adjacent to petrochemical industrial counties (PICs). Directly age-adjusted mortality rates for cancer during 1982-1991 among 16 counties characterized by a heavy concentration of petrochemical industries were compared to rates among 16 matched counties with similar concentration of nonpetrochemical manufacturing industries, urbanization level, and demographic characteristics. An excess rate for liver cancer among males was found in the so-called PICs. The correlation could not be explained by confounding variables such as urbanization, socioeconomic class, or employment in nonpetrochemical industries. No other increased cancer risks were found to be associated with residence near petrochemical industries.