The way in which children are questioned in forensic contexts can impact on the accuracy of their responses. Past studies have shown that children were often questioned in the New Zealand courts in ways that profoundly contradict best practice. This study analyses the questions posed to 18 child witnesses by forensic interviewers, prosecutors, and defence lawyers during criminal trials held in New Zealand courts in 2008. The results suggest that, as was found in earlier studies, many of the questions posed to children during cross-examination in particular were inconsistent with best practice in terms of eliciting full and accurate information from children. Indeed, the heavy reliance on closed, leading and complex questions, along with other common practices, casts doubt on the forensic safety of cross-examination, calling into question its fairness for children. The implications of these results for policy and practice are explored.