A selective review of the current methods of pain measurement and validation (psychophysical methods, verbal and analogical scales, psychological tests) is presented with emphasis on patient selection for surgical pain relief, and analysis of outcome. The identification of homogeneous groups of patients with clinical and research objectives is prevented by the lack of a reliable pain scale, based on the assessment of objective and comprehensive parameters. This obstacle seems to be inherent to the complex nature of human pain experience. Psychiatric examination has proved important to elucidate the operative indications, particularly in cases of non-malignant obscure neuralgias. The importance of separate validation of the pain compliant and the psychiatric assessment is stressed. A critical comment is made on Hitchcock's pain scale and Lindqvist's psychiatric classification of candidates for surgery.