Health technology assessments incorporating measurement of quality of life are becoming increasingly common. In the evaluation of magnetic resonance imaging (MRI), the most widely used quality of life instrument has been the Rosser classification. The theoretical basis for incorporating such a measure in the evaluation of imaging technology is considered and several studies using this instrument are reviewed. Despite the obvious technical advances of MRI and its influence on clinical management, it has proved difficult to demonstrate that the use of this expensive new imaging technique contributes to improvement in health-related quality of life. Some reasons why the expected improvement has not been demonstrated are discussed, as are the quality of life instruments which may be appropriate in this setting.