The main aims of this study were to assess memory of pain and pain-related affect in headache sufferers and to compare the pain memories of individuals who suffer from migraines and those who experience non-migraine headaches. A total of 313 women participated in the study, 86 of whom had a diagnosis of migraine headache. The women rated their state anxiety about having a headache, the intensity and unpleasantness of headache pain, and their emotions while actually experiencing a headache. Either three or six months later, the participants were asked to recall their state anxiety, the intensity and unpleasantness of pain, and the emotions they had felt. Regardless of the length of recall delay or migraine diagnosis, participants accurately remembered both pain intensity and unpleasantness. Together, recalled anxiety, experienced pain and recalled positive affect were the most important predictors of memories for headache, accounting for 41% and 37% of the total variance in recalled pain intensity and unpleasantness, respectively. However, participants overestimated recalled positive and negative affect. The effect of recalled affect on memory of headache, together with the overestimation of recalled affect, suggests that although memory of headache is accurate, it is influenced by distorted memories of affect.