BACKGROUND Despite the problems involved in treating depression and concomitant medical disease, there are virtually no longitudinal studies on drug utilisation among depressed patients. METHODS Use of prescription drugs among all first-time users of antidepressants in a defined population five years before and six years after the index (first) treatment was compared to a referent group without antidepressant treatment. The generalised estimating equations (GEE) method was used for analysis. RESULTS The antidepressant-treated group used considerably more non-psychotropic drugs during the whole study period than the referent group. They also used more psychotropic drugs, a use which increased in connection with the initiation of antidepressant treatment, and stayed high for a further five years. CONCLUSIONS The high use of prescription drugs indicated widespread somatic and psychiatric health problems during the whole study period. Antidepressant-treated patients are at risk for drug interactions and adverse effects, and would benefit from a closer collaboration between psychiatry and medicine.