The methods literature on research with children recognises the challenges of negotiating informed consent with this group. Special "child-friendly" techniques are advocated to overcome these challenges. We argue that, upon closer inspection, research with children foregrounds more fundamental problems with informed consent that are not easily resolved. Drawing from three ethical texts commonly consulted in our own research fields, we highlight problems of information, understanding, authority, capacity and voluntarity. We conclude that informed consent is more problematic than is generally admitted, and that researchers would benefit from more openly acknowledging its limitations.