Previous researchers found that young children will try to answer nonsensical questions. In Expt 1, 5- to 8-year-olds were asked sensible and nonsensical questions. Half of each type were ‘closed’ questions (which required a yes/no response), and half were ‘open’ questions (which could be answered in several ways). Three weeks later the same children were asked to judge if the questions were sensible or silly. Children answered all the sensible questions appropriately, and only attempted to answer a small proportion of the nonsensical open questions. However, they did try to answer three-quarters of the nonsensical closed questions. Nonetheless, children were nearly always correct in judging which questions were sensible and which were nonsensical. In Expt 1 all the closed nonsensical questions were also ones that required a comparison between two items. In Expt 2 we compared children's responses to nonsensical open and closed questions when half of each type were comparative and half were non-comparative. Children attempted to answer nonsensical closed questions irrespective of whether or not they included a comparison. However, few children attempted to answer nonsensical open questions. We discuss the implications of these results for questioning children and in the context of children's eyewitness testimony.