In formal police interviews, interviewers may have institutionally ordered reasons for following up even seemingly fully co-operative replies with questions that insinuate that the interviewee is in fact (consciously or unconsciously) being uncooperative. From a sample of over 100 UK interviews with suspects apprehended for minor crimes, and 19 interviews with witnesses claiming sexual assault, the authors distinguish and explore follow-up questions which do not suppose that interviewees’ seemingly ‘normal’ replies respect the Gricean maxims of quantity, quality, relevance or manner. Three institutional motivations working to over-ride the normal communicative contract are distinguished: to ‘get the facts straight’; to prepare for later challenges; and pursue a description of events that more evidently categorises the alleged perpetrators’ behaviour as criminal.