In epidemiologic research on chronic pain, differentiation of recurrent, persistent and disabling pain states is critical in the investigation of burden, natural history, effective intervention and causal processes. We report population-based data concerning the development and evaluation of a graded classification of pain status. In a probability sample of 1016 health maintenance organization enrollees, recurrent or persistent pain was observed in 45%; severe and persistent pain in 8%; severe and persistent pain with 7 or more days of pain-related activity limitation in 2.7%; and severe, persistent pain with activity limitation and 3 or more indicators of pain dysfunction in 1.0% of the population sample. Graded chronic pain status was associated with psychological impairment, unfavorable appraisal of health status, and frequency of use of pain medications and health care. The presence of severe and persistent pain increased the likelihood of multiple indicators of pain dysfunction, but there was considerable heterogeneity in pain dysfunction among persons with comparable pain experience. Our data suggest grading chronic pain in terms of 3 axes: time (persistence); severity; and impact (disability and dysfunctional illness behaviors).