The study examined the associations between several sociodemographic and psychosocial variables and unexplained chronic fatigue in the community before and after adjustment for psychiatric morbidity and determined the prevalence of fatigue and rate of disability resulting from fatigue in the general population. The study is a secondary analysis of 1993 data from a household survey of psychiatric morbidity conducted by the Office for Population Censuses and Surveys in Great Britain. The survey included 12,730 subjects age 16-64 years. Unexplained chronic fatigue was used as the dependent variable in a logistic regression analysis, with various sociodemographic and psychosocial variables and psychiatric morbidity as the independent variables. Psychiatric morbidity was assessed by using the Revised Clinical Interview Schedule. Fatigue was measured by using the fatigue section of the Revised Clinical Interview Schedule. A total of 10,108 subjects agreed to cooperate (79.4% participation rate). The prevalence of unexplained chronic fatigue was 9%. Subjects with psychiatric morbidity had higher rates of fatigue. Adjustment for psychiatric morbidity had a minor effect on the associations between sociodemographic factors and chronic fatigue. After adjustment, older subjects, women, and couples with children had higher rates of fatigue. Single subjects, widowed subjects, adults living with parents, and economically inactive subjects had lower rates of fatigue. Fatigue was associated with considerable disability, but the association between fatigue and psychiatric morbidity explained most of this disability. Unexplained chronic fatigue is a common condition, strongly associated with psychiatric morbidity. The close relationship between fatigue and psychiatric morbidity should not obscure the possibility of differences as well as similarities in their aetiologies.