Aims: In Sweden, the Social Insurance Offices each year refer long-term sickness absentees to comprehensive investigations to clarify medical conditions. However, there is a lack of scientific knowledge about these patients and their morbidity. The aim was to characterize a population of these sickness absentees regarding prevalence of somatic and psychiatric diagnoses and possible associations with sociodemographic, lifestyle, and health characteristics. Methods: A cross-sectional study was made up of 635 sickness absentees below the age of 64, who the local Social Insurance Offices in Stockholm County, Sweden, referred to a special unit for multidisciplinary investigation. Data was obtained from questionnaires and medical records. The patients were examined by board certified specialists in psychiatry, orthopaedic surgery, and rehabilitation medicine. Relative risks were estimated by use of modified Poisson regression to assess the associations between characteristics and diagnose outcomes. Results: About 80% of the patients had more than one diagnosis. The vast majority had a psychiatric diagnosis, and approximately 55% had that in combination with at least one somatic diagnosis. An increased risk for being given a psychiatric diagnosis was found for men and unemployed people. In addition, lack of social life and friends and self-reported mental health problems were associated with psychiatric diagnoses but also among those who were given somatic diagnoses in combination with psychiatric diagnoses. Increased risks for somatic diagnoses were found for women and for patients with a higher education. Conclusions: Long-term sickness absentees referred to multidisciplinary investigations display high co-morbidity of psychiatric and somatic diagnoses and are a heterogeneous group with diverse sociodemographic and medical characteristics.