The possibility of a cancer risk associated with fluoridation of public water supplies is re-examined using mortality data for 35 US cities, 20 with fluoridated water supplies and 15 with non-fluoridated water. Crude cancer death rates, and mortality ratios standardized for age, sex, and ethnic group are examined, using four alternative sets of standard rates and three different combinations of pericensal years' deaths. Changes in cancer mortality before and after fluoridation in the fluoridated cities are compared with changes in the non-fluoridated cities over the same time period. In none of the analyses have differences in mortality trends been found that could not be due to chance alone. Thus, these results do not support the suggestion of an association between fluoridation of water supplies and cancer mortality.