Medicines have the potential to cause harm, particularly when adolescents do not follow recommendations for use. In addition, medicine use in adolescence has been shown to track into adulthood. There is therefore a strong rationale to monitor changes in adolescent medicine use over time and understand the mechanisms behind these changesData from the 1998, 2006 and 2010 Scotland Health Behaviour in School-aged Children Survey were modelled using multilevel logistic regression, modelling medicine use for: headache, stomachache, sleeping difficulties and nervousness, as well as a combined medicine use measure. Models adjusted for year and frequency of health complaints to measure trends in medicine use, and an interaction term to measure the relationship between medicine use and health complaint frequency.Medicine use reduced between 1998 and 2010. Hownever having the majority of the reduction was observed between 1998 and 2006 for all five outcomes. Adjustment for health complaint frequency only explained some of this reduction. When an interaction term was added between year and health complaint frequency this was significant for boys' medicine use, suggesting that health complaint frequency became a better predictor of medicine use with time. Medicine use for stomachache among girls increased over time, and this increase became more pronounced after adjustment for stomachache frequency.Changes in health complaint frequency were only partly associated with reductions in medicine use between 1998 and 2010. Further monitoring of medicine use is recommended, particularly the use of medicine for sleep difficulties, and for stomachache among girls.