We developed a 12-item questionnaire for completion by patients having shoulder operations other than stabilisation. A prospective study of 111 patients was undertaken before operation and at follow-up six months later. Each patient completed the new questionnaire and the SF36 form. Some filled in the Stanford Health Assessment Questionnaire (HAQ). An orthopaedic surgeon assessed the Constant shoulder score. The single score derived from the questionnaire had a high internal consistency. Reproducibility, examined by test-retest reliability, was found to be satisfactory. The validity of the questionnaire was established by obtaining significant correlations in the expected direction with the Constant score and the relevant scales of the SF36 and the HAQ. Sensitivity to change was assessed by analysing the differences between the preoperative scores and those at follow-up. Changes in scores were compared with the patients' responses to postoperative questions about their condition. The standardised effect size for the new questionnaire compared favourably with that for the SF36 and the HAQ. The new questionnaire was the most efficient in distinguishing patients who said that their shoulder was much better from all other patients. The shoulder questionnaire provides a measure of outcome for shoulder operations which is short, practical, reliable, valid and sensitive to clinically important changes.