A procedure for the evaluation of fluctuations in perceived pain intensity among chronic pain patients is described, and its psychometric properties and clinical and heuristic utility are examined. A heterogeneous sample of 97 chronic pain patients recorded 2 weeks of hourly self-monitored pain intensity (SMPI), completed a structured interview and several questionnaires, and established behavioral goals prior to participation in a pain rehabilitation program. Three variables were derived from the SMPI data: mean SMPI, variability, and number of missing observations. A series of analyses supported both the test-retest reliability and the concurrent validity of SMPI. Significant correlations with measures of depression, anxiety, marital satisfaction, perceived life interference, and activity levels were interpreted as support for the conceptual validity of SMPI within a cognitive-behavioral perspective. The utility of SMPI in predicting rehabilitation outcomes was also demonstrated.