Aim: The aim was to investigate the association between psychosomatic problems and lifestyle patterns of alcohol use and physical exercise in adolescence. Subject and Methods: Cross-sectional questionnaire data from 8 years of investigations (1988–2011) were used. The sample included over 20,000 adolescents (50 % girls), aged 15–16. Adolescents were divided into three groups depending on psychosomatic problem load: 10th, 11th–89th and 90th percentiles on the psychosomatic problems scale (PSP). A composite measure of alcohol use (frequent/rarely/never) and physical exercise (regular/little/never) was created, resulting in nine combinations of health-related behaviours, and used as dependent variable. Results: Multinomial regression analysis showed that those over the 90th percentile of the PSP had nearly 12 times higher odds of regular alcohol use combined with no exercise compared with those who had the least symptoms. The former subgroup also had higher odds of belonging to all suboptimal lifestyle categories. Those in the 11th–89th percentiles had increased, albeit smaller, odds of belonging to all lifestyle categories. Descriptive data analysis indicated gender differences and changes over time in the strength of the association between psychosomatic problems and health-related behaviours but the regression analysis did not show any statistically significant interactions. The proportion of adolescents engaging in the unhealthiest lifestyle had decreased over time while reports of psychosomatic problems had increased. Conclusion: Adolescents with the greatest psychosomatic symptom load were the most likely to engage in unhealthy lifestyles. This may mean multiplied risks of future psychopathology and warrants longitudinal investigations