Background. Despite the clinical and economic importance of depression, resource utilization and cost in clinical practice is not well documented. The aim of this study was to investigate medical resource consumption, productivity loss and costs associated with patients treated with antidepressants for depression in primary care. Methods. 447 patients were enrolled at 56 Swedish primary care centres to this naturalistic longitudinal observational study. Patients over 18 years with depressive symptoms, and who initiated a new antidepressant therapy were included in the study. Data on patients' socio-demographics, daily activity and quality-of-life (EQ-5D) were collected using questionnaires completed during outpatient GP visit for a follow-up period of approximately 6 months. Results. Based on a complete sample of 398 patients, the total annual cost per patient was estimated at $13 400 (SEK 12 300-$15 100) in 2005 prices. Direct costs were estimated at $4 800 ($4 300-$5 400), constituting 35% of the total annual cost per patient. Among direct costs, the cost for medical visits was the largest single cost item, representing about 18%. The costs for antidepressants represented only 4% of the total costs. The indirect costs, i.e. productivity loss due to lost working time, were estimated at $9 000 per patient ($7 600-10 200), or 65% of the total annual costs per patient. No demographic variables were significantly associated with cost of depression. Direct and indirect costs were however correlated positively with achievement of clinical remission. The presence of sick-leave during follow-up was moreover associated with 1.8 times higher costs. Conclusions. The burden of depression to society is high, both in terms of direct treatments costs and indirect costs for sickness absence and early retirement. Because of the high indirect cost per patient, there seems to be a particular need for therapies that have the potential to reduce absenteeism.