Patients' recall of their recent pain is commonly assessed by both researchers and clinicians. However, concerns have been raised about differences between recalled pain and the average of real-time recordings of pain taken over the same period, in part because of the possibility that memory processes affect how retrospective pain is reported. It was hypothesized that memory processes affect the accuracy of recall, such that those with higher versus lower variability of real-time pain will recall pain at higher levels, relative to their average momentary pain. Sixty-eight chronic pain patients with rheumatologic conditions reported their pain several times a day for 2 weeks and also recalled their weekly pain at the end of each of those 2 weeks. The hypothesis was confirmed and it was concluded that variability of real-time pain affects the recall of pain.