Abstract
This paper discusses and attempts to estimate the very small numbers of women attending the UK Breast Screening Programme for whom the risk of cancer induction may exceed the probability of cancer detection. It updates a previous paper on the same topic. Variations in breast dose between individuals, due to differences in breast size and in numbers of views and films taken, are considered and revised. New data on cancer induction and its variation with age at exposure have been employed. The overall effect of these changes is generally to improve the balance of benefit against risk compared with the previous paper referred to, the very few exceptions being categories where the numbers of women in question remain of the order of one in a million. The implications for certain alternative screening schedules and for some current trials are also discussed, the conclusions being again reasonably reassuring.