Abstract
Registration of pain must take into consideration the private nature of the pain experience while providing researcher and practitioner with measures which obtain satisfactory levels of compliance, reliability and sensitivity. A three-component model of pain based on social learning theory, comprised of physiological, cognitive/affective and overt behavioural dimensions has been shown to be a valuable tool in monitoring pain. Applications of the model show that it has satisfactory methodological characteristics when used in a study with continuous self-recording by patients, that it can accommodate discrepancies in physiological and affective dimensions and that it offers a variety of sensitive measures for evaluating treatment of pain.