The inverse of a difference in probabilities, called the ‘number needed to treat’, has been promoted in the medical literature as a good way to present the results of modelling binary outcomes. The usual context is randomized controlled trials and meta-analyses. In this paper we discuss the claims that have been made about this statistic, and the problems associated with it. Methods which have been proposed for confidence intervals are evaluated and shown to be erroneous. We suggest that the difference in probabilities, the ‘absolute risk reduction’, is preferable to the number needed to treat, for both theoretical and practical reasons.