Early geographical studies of cancer in areas that have naturally-occurring fluoride at different levels gave no indication of an effect on cancer rates associated with higher intakes of fluoride. Following widespread fluoridation to improve dental health in the United States and Britain, non-epidemiologists presented analyses of cancer data which they claimed demonstrated such an effect. However, subsequent large-scale comparisons of cancer rates in fluoridated and non-fluoridated areas for successive periods following fluoridation have not indicated any increase, either for all cancer or for malignancies across the range of individual sites. Studies undertaken specifically to examine the claims of the non-epidemiologists have, time-and-again, shown that, with the use of accurate data and correct statistical methods, the purported effects cease to be apparent. Details of the earlier evidence and claims are given in the 'Report of a Working Party on the Fluoridation of Water and Cancer' by Professor George Knox (1985) and of more recent analyses in Hoover et al. (1991a; 1991b; unpublished internal US PHS Memo, 1993). The present paper gives a brief overview of the evidence that fluoride in drinking water has not been shown to cause an increase in the risk of developing cancer and of the errors in the analyses that purport to show such an increase.