This study attempted to predict postoperative pain from preoperative level of anxiety and the amount of information patients possessed regarding their surgery. Pain was assessed via the McGill Pain Questionnaire (MPQ) and a measure of pain complaints--number of analgesics taken. High levels of state anxiety and a high degree of information predicted the Present Pain Intensity measured of the MPQ, but did not predict the Pain Rating Index portion of the MPQ. The number of analgesics taken was predicted from the amount of information but not the level of presurgical anxiety. Biographical variables were unrelated to postoperative pain. The results were discussed in terms of State-Trait Anxiety theory, Janis' curvilinear prediction model and a contextual perspective of information imparting.