Factors influencing natural history and clinical course of pain in temporomandibular disorders (TMD) are largely unknown. Physical, psychological and behavioral data from a population-based epidemiologic study of TMD were examined in 234 cases of persons reporting TMD pain. The cases were assigned to one of five pain pattern groups based on changes in average TMD pain from baseline to 5-year follow-up: (i) remitted (49% of the sample), (ii) high-improvement (14%), (iii) low-improvement (9%), (iv) same (13%), and (v) worse (16%). For each pain change group, an ANOVA-derived pattern analysis was performed to assess whether the pattern of change in each of seven physical and three psychological variables was congruent or dissimilar to the pattern of change in average pain intensity. For none of the physical or psychological variables was the change over time completely congruent with the changes in pain. Changes in ambient average TMD pain were most closely related to those clinical variables whose assessment is influenced by pain or other self-reported symptoms (e.g., number of muscle sites painful to examiner palpation), while the amount of pain change was less closely related to changes in clinical variables, such as joint sounds, where assessment is not dependent on subjective report. The three psychological variables, anxiety, depression, and somatization, displayed similar change patterns, but these patterns were distinctly different from those of the physical variables in that the remitted pain group was at the population mean at baseline for these psychological variables and remained there; significant improvement in psychological status was observed only in the pain group showing high improvement. The other three pain change groups exhibited elevated psychological distress scores at both baseline and 5 years. These results indicate that although the relationships among the course of pain, of physical variables, and of psychological variables are complicated, the 5-year outcome in pain is largely independent of readily discernible changes in clinical signs.