The importance of measuring the Quality of Life (QL) has become more and more apparent during the past 10 years. Traditionally, QL studies have investigated functional status and treatment side effects. In recent years more comprehensive instruments have been constructed to assess the QL of cancer patients. Most QL instruments take a certain degree of physical and psychological functioning into account as well as the level of activity and a global evaluation of life. The reliability and validity of these instruments are reviewed in this paper. Reliability has been well documented and seems satisfactory for most instruments. Only a few authors have discussed content validity. Predictive validity has also been established in most instances and the results from various studies are promising, however, the choice of criteria is seldom taken into account. It also seems difficult to draw conclusions about construct validity. In our view, the main problem is the lack of definitions and the absence of a theoretical framework. Therefore, some assumptions underlying QL research in cancer patients have been elucidated.