Abstract: This investigation assessed the relationships among recalled, expected, and actual muscle pain intensity and unpleasantness during a period of 48 hours. We hypothesized that (1) specific expectations of pain after 24 hours would account for a significant amount of variance in actual pain, (2) recalled pain from the most recent episode of naturally occurring muscle pain would be significantly associated with expected pain, and (3) the accuracy of expectations (ie, the difference between expected and actual pain) would increase across time as the onset of muscle pain occurred. Ninety-five students completed 3 sessions. In Session 1, recalled muscle pain and expected muscle pain in the next 24 hours were collected before exercise. In Sessions 2 and 3, muscle pain during movement and expected pain in the next 24 hours were collected. Recalled muscle pain was associated with expectations at baseline, r values [equals] .26 to .47, P [lt ] .05. The accuracy of expected intensity increased during the study, t [equals] 3.20, P [lt ] .01, and the accuracy of expected unpleasantness was associated with change in expected unpleasantness, r values [equals] [minus ].28, P [lt ] .01. The amount of variance in actual intensity and unpleasantness accounted for by expectations increased up to 55% and 52%, respectively, during the study. Expected unpleasantness did not account for variance in actual intensity. Expected intensity accounted for 3% of the variance in actual unpleasantness, but only in the second 24-hour period. Thus, our hypotheses were generally supported, but unanticipated findings regarding changes across time in the relationships among recalled, expected, and actual muscle pain were also detected. [copy ] 2003 by the American Pain Society