This paper explores challenges arising from research (Daniels & Macnab, 2004), which sought to identify young people aged 14–16 years who were emotionally vulnerable and not in receipt of educational provision within two local UK authorities. Two particular challenges are identified in this research. The first was the challenge of defining the term used to identify the young people. Second, the research faced an almost insurmountable challenge in finding and obtaining data on a sample for which the outcomes of the research may pose a threat to those holding the data. Discussion focuses upon possible ways in which research can address these issues.