The principle of Einstein's theory of special relativity is that an observer of an apparently moving body cannot be sure if the body really has moved, if he/she has moved or if both events have occurred. Although Einstein was discussing physical events, a similar hypothesis may apply to quality of life. When using quality of life instruments, one presumes that the point of reference (the observer in Einstein's terms) does not move, i.e. that an individual's attitude towards a particular construct will remain stable. Otherwise, changes in response to particular variables cannot be interpreted. However, attitudes are not constant: they vary with time and experience and are modified by such psychological phenomena as adaptation, coping, expectancy, optimism, self-control and self-concept. For example, eating problems may be extremely important at one point in a person's life. However, when oral discomfort has been diagnosed as cancer and treated with surgery or radiation, the same individual may "objectively" demonstrate more problems when eating, but report them as less because they have now become relatively unimportant. Furthermore, paradoxical reports that some groups of ill individuals rate their quality of life higher than do "healthy" persons raise similar questions concerning between-group point of reference differences. Investigators in the fields of organisational management, education and psychology have developed techniques such as "then ratings", saliency indicators and individualised questionnaires in attempts to quantify within-subject variability and between-group differences pertaining to point of reference. We suggest that similar methods may help us to measure change in the impact of the different items of quality of life instruments. In this paper, we will describe the theories of change associated with quality of life measurement. In addition, we will present evidence suggesting that the point of reference does change, the reasons for this and possible solutions to the problem.