Since pro-arrest policies in domestic violence became popular in the United States in the 1980s, numerous western countries have followed suit. In most cases, research has shown that implementation of the policies has fallen short of expectations, with arrest rates that are surprisingly low. In New Zealand, pro–arrest strategies have been employed since 1987 and results have been similar. This article argues that one of the reasons for noncompliance in New Zealand (and probably elsewhere), is that the complexities of domestic violence situations make pro–arrest difficult to apply in practice. Moreover, in order to protect themselves from official criticism for deviating from policy, in this study frontline police sometimes filed incomplete or inaccurate incident reports. This made it hard to determine exactly how well the policy was being implemented and whether or not it was working.