The present study was conducted to determine whether demographic variables, medical status variables, and psychological measures at pretreatment were related to pain reduction immediately following behavioral treatment for headache and at a 6-month follow-up. The study sample consisted of 156 subjects, who were selected for participation in a behavioral outcome study on the efficacy of autogenic training and cognitive self-hypnosis training. A Headache Index based on pain diaries constituted the main outcome measure. Psychological measures included the Symptom Checklist-90, Dutch Personality Questionnaire, Coping Strategy Questionnaire, Multidimensional Locus of Pain Control Questionnaire, and treatment expectations. Subjects who expected more pain reduction at pretreatment achieved a lower level of pain at posttreatment, independent of pretreatment pain levels. None of the other pretreatment variables were related with pain reduction at posttreatment or at the follow-up. Finally, at the 6-month follow-up, 43 subjects were classified as responders (more than 50% pain reduction) and 113 as nonresponders (less than 50% pain reduction and dropouts). At pretreatment, the responders perceived more pain control than the nonresponders. None of the other pretreatment differences between responders and nonresponders proved to be significant. The main conclusion that could be drawn from this study was that pain reduction, in the short- and long-term, cannot be predicted with any accuracy by demographic and medical status variables or scores for psychological distress, personality traits, coping strategy use, and pain appraisals.