In this paper we present a case study of the gendering of evidence in family policy reform. We examine the characterisation of data as legitimate or illegitimate in a recent Australian Inquiry into child support and custody issues. Using critical discourse analysis, we examine how Inquiry committee members interpreted personal anecdotes and social scientific data presented by witnesses. Data were characterised as legitimate when confirming an existing stock story of fathers' disadvantage in an unfair child support system; these data were treated as evidence of a widespread social problem. Data that challenged the stock story were rejected as offering an inappropriate basis for policy reform. We conclude that in these reforms, the type of data, be it social scientific or anecdotal, was not as important as its alignment with a priori understandings of the nature of the world and the policy problems to be solved.