The association between active musical engagement (as leisure activity or professionally) and mental health is still unclear, with earlier studies reporting contrasting findings. Here we tested whether musical engagement predicts (1) a diagnosis of depression, anxiety, schizophrenia, bipolar or stress-related disorders based on nationwide patient registers or (2) self-reported depressive, burnout and schizotypal symptoms in 10,776 Swedish twins. Information was available on the years individuals played an instrument, including their start and stop date if applicable, and their level of achievement. Survival analyses were used to test the effect of musical engagement on the incidence of psychiatric disorders. Regression analyses were applied for self-reported psychiatric symptoms. Additionally, we conducted co-twin control analyses to further explore the association while controlling for genetic and shared environmental confounding. Results showed that overall individuals playing a musical instrument (independent of their musical achievement) may have a somewhat increased risk for mental health problems, though only significant for self-reported mental health measures. When controlling for familial liability associations diminished, suggesting that the association is likely not due to a causal negative effect of playing music, but rather to shared underlying environmental or genetic factors influencing both musicianship and mental health problems.