Assessing self-harm through brief multiple choice items is simple and less invasive than more detailed methods of assessment. However, there is currently little validation for brief methods of self-harm assessment. This study evaluates the extent to which adolescents' perceptions of self-harm agree with definitions in the literature, and what level of question detail produces optimal concordance rates. Two hundred and thirty-three (69% female) first year university students aged 17-21 completed a self-harm coding task; we created three levels of question detail and randomly allocated participants to three study groups: "brief", "low detail", and "high detail". The present findings suggest that that adolescents' perceptions of self-harm are generally concordant with a consensus definition of self-harm. Low level of detail in the question produced greatest accuracy in responses; adolescents who demonstrated adequate task understanding were able to correctly identify 94% of examples of self-harm behaviour and 86% of examples of behaviour that were not self-harm. We identified lower concordance rates for eating disordered behaviour and recreational petrol sniffing. This indicates that adolescents perceive these behaviours to be self-harm, in contrast to the reference definition we utilised. Overall, this study provides support for using a brief assessment of self-harm where minimal detail regarding self-harm behaviour is required.