As most young people who self-harm do not present to clinical services for their self-harm, capturing accurate data on self-harm, beyond that available from healthcare records, is crucial for research and prevention efforts. Moreover, self-harm thoughts and behaviours are dynamic, varying over minutes and hours, yet traditional, retrospective self-report measures — commonly used in self-harm research — miss this variability and can be subject to recall bias. To understand the full picture of self-harm thoughts and behaviours in young people, we need to take research on self-harm out of the lab and the clinic and into young people’s everyday lives, for example, by using experience sampling methods (ESM), a type of structured, electronic diary technique. In the SIGMA study, we used school-based recruitment to gather data from N=1235 adolescents (aged 12, 14, or 16 at onboarding) from the general adolescent population in Flanders (Northern Belgium), using a mixture of self-report questionnaires, ESM, and wearables. Lifetime history of self-harm thoughts and behaviours were assessed at baseline, and current self-harm thoughts and behaviours were assessed using ESM, 10 x per day for six days. 59% of adolescents reported no lifetime history of self-harm thoughts or behaviours, 20% reported self-harm thoughts, and 21% reported self-harm behaviours. Adolescents also reported self-harm thoughts and behaviours during the six-day ESM period. We found no differences between adolescents with and without self-harm thoughts and behaviours in ESM compliance or beep disturbance, however, we found that when adolescents’ self-harm thoughts were more intense, they also reported ESM as more disruptive. Our research shows that using ESM to study self-harm in school-going adolescents is feasible and acceptable, and can provide crucial new insights into the dynamic nature of self-harm thoughts and behaviours. Results also suggest that ‘static’ conceptions of vulnerability for research participation based only on lifetime history are not supported, and a more dynamic approach to vulnerability is warranted.