OBJECTIVE: Studies have shown a greater use of medical than mental health services in patients with somatoform disorders. However, not many studies are based on structured interviews and include the entire somatoform spectrum of diagnoses. We conducted a register-based case-control study to investigate medical care use prior to and three years after diagnosis in patients with somatoform disorders. METHODS: We included 380 patients with somatoform diagnoses (SCID-NP for DSM-IIIR) in a case-control study and compared them with 174 patients with anxiety disorders and 5540 controls from the background population. Data from the Danish National Registers were used to assess health care use in both primary and secondary care. RESULTS: Somatoform patients incurred 2.11 (2.09-2.12) times the primary care visits of controls. They had 3.12 (3.08-3.16) times as many somatic bed-days than controls and 3.94 (3.91-3.97) as many psychiatric bed-days. Primary care use remained stable 3 years after diagnosis (p = 0.14) and the award of disability pension (p = 0.82). However, the number of somatic admissions decreased from 5.64 to 2.76 (p = 0.0004) 3 years after diagnosis. Somatization had an independent effect on health care use when controlling for psychiatric comorbidity. CONCLUSIONS: Patients with somatoform disorders make significantly greater use of health care services than do controls and patients with anxiety. Somatoform patients made more use of psychiatric services than expected. The use of somatic health care was independent of psychiatric comorbidity. Primary care use and disability pension award were not influenced by proper diagnosing of somatoform disorders whereas number of somatic admissions were halved.