OBJECTIVE: Chronic fatigue syndrome (CFS) case definitions agree that fatigue must be unexplained, debilitating and present for at least 6 months, but they differ over accompanying symptoms. Our objective was to compare the 1994 CFS case-defining symptoms with those identified by factor analysis. METHODS: We surveyed the Wichita population and measured the occurrence of 21 symptoms in 1391 chronically fatigued subjects who did not report fatigue-associated medical or psychiatric conditions. We used factor analyses to identify symptom dimensions of fatigue and cluster analysis to assign subjects to subgroups. RESULTS: Forty-three subjects had CFS. We confirmed three factors: musculoskeletal, infection and cognition-mood-sleep, essentially defined by CFS symptoms. Although factor scores were higher among CFS subjects, CFS and non-CFS distributions overlapped substantially. Three clusters also showed overlap between CFS and non-CFS subjects. CONCLUSION: CFS symptomatology is a multidimensional phenomenon overlapping with other unexplained fatiguing syndromes and this must be considered in CFS research.