As part of the recent 'therapeutic turn' in education, schools are now commonly seen as a place for mental health guidance and support. This often involves "interventions"--special curricula of lessons or activities (e.g. counselling sessions), which aim to either prevent mental health problems or manage those that have already started. Running these interventions in schools makes good sense: rates of mental health problems in young people are rising, and large numbers can be reached in this setting. However, evidence for the effectiveness of such interventions has been mixed. One way to improve how helpful and useful they are, we argue here, would be to ask young people themselves what they think about these programmes. This involves collecting "qualitative" data: gathering in-depth information about young people's experiences and opinions, rather than relying solely on numerical data, such as rating scales. The small number of existing published qualitative studies in this area show that many young people do find these interventions helpful, but there are issues that warrant careful attention. For example, some young people can feel worried or vulnerable during classroom-based exercises, and others don't see how the interventions are relevant for their own lives. Here, we explore this literature and recommend two avenues for future work: ask more young people what they think of existing interventions, and get them involved in the design of new ones. Together, this will put young people's voices at the heart of school-based mental health interventions.