A cross-sectional descriptive design was utilized to explore expected, experienced and recalled pain, and anxiety. Subjects were 138 children (5-17 yr) who were attending an outpatient laboratory for venipuncture. They reported state anxiety and expected visual analogue scale pain prior to having venipuncture. Following venipuncture, they reported sensory and affective pain. Two months afterward, subjects were contacted and asked to recall sensory and affective pain from the venipuncture. Analysis of data determined that children's recall of pain was quite good although better for affective pain than sensory pain. Prediction accuracy for pain was not good. From analysis it was determined that overestimation of pain was related to anxiety but not to greater experienced pain. Four patterns of responses were identified when prediction and recall accuracy were further examined. These patterns were designated as realism, irrelevance, over-reaction and denial. It was suggested that these patterns of response may provide clues about avoidance and coping behavior for children's pain. In general, the results of this study did not support previous research. It was demonstrated that various statistical techniques which have been employed in research on this topic have influenced the interpretation of pain prediction and recall data.