Somatization is a widespread problem in health care. We estimated the occurrence of Somatization Disorder (SD) using three different case-finding methods in a general population cohort. The sample consists of 1,598 subjects born in 1966. The case-finding methods according to the DSM-III-R criteria for SD were: 1) Finnish Hospital Discharge Register (FHDR) data, 2) analysis of the patient records in public outpatient care 1982-1997, and 3) Structured Clinical Interview for DSM-III-R (SCID) for 321 selected cases. The prevalence of SD was 1.1% (N = 18), giving a female-to-male ratio of 5:1. All cases were found among the public outpatient care records. No cases appeared in the FHDR or were recognized in the psychiatric interview. The lifetime prevalence of SD was comparable with previous western population studies. Methodologically, information from outpatient records may be more sensitive in detecting SD than hospital diagnosis or even psychiatric interview. Clinically we stress the importance of recognizing these cases by liaison psychiatrists especially because SD has been recognized as being difficult to treat among somatic and primary health service providers and because some promising treatment alternatives such as cognitive-behavioral therapy and antidepressants have emerged for SD patients.