OBJECTIVES: To assess the reliability, validity, and responsiveness of a new quality of life measure, the patient generated index (PGI) of quality of life, in patients with four common clinical conditions. DESIGN: Prospective one year follow up study. SETTING: Outpatient departments and four general practices in Grampian, Scotland. SUBJECTS: 1746 patients consulting a general practitioner in one of four practices, or referred to outpatients from all Grampian practices over a four month period, with low back pain, menorrhagia, suspected peptic ulcer, and varicose veins. MAIN OUTCOME MEASURES: Postal questionnaire including the PGI, SF-36 health survey, and clinically derived condition specific measures of disease severity. RESULTS: Test-retest reliability was satisfactory for group comparisons (intraclass correlation coefficient 0.65). Validity was confirmed by the observed association of the PGI with the SF-36, condition specific instruments, and sociodemographic variables. For low back pain, the PGI and the SF-36 pain scale were found to be most responsive to clinical change. For patients with menorrhagia and suspected peptic ulcer, only the condition specific instruments detected larger changes than the PGI. CONCLUSIONS: It is possible to develop a patient generated index of quality of life that not only assesses the extent to which patients' expectations are matched by reality but also satisfies criteria of reliability and responsiveness to change. Further work is required to make the PGI more acceptable and meaningful to patients, but it is believed that it offers an exciting new approach to the evaluation of medical care.