This thesis examines the use quality adjusted life years (QALY'S) in the allocation of health care resources. It is divided into three broad sections. The first section discusses how health status measurement techniques can be used to derive the utility values incorporated into QALY'S. The second section uses one health status measurement instrument, the Rosser-Kind Classification of Illness States, to estimate the QALY'S gained by patients who have undergone hip and knee joint replacement surgery. It is shown that the Rosser-Kind Classification of Illness States is as effective in measuring the health-related quality of life of these patients as more detailed questionnaires. In addition, it is found that further research is required before any generalisations concerning the acceptability of retrospective data can be made. A third important result is that there are significant improvements in health- related quality of life following both types of surgery, with the highest Rosser- Kind rating scores achieved after the first year following knee replacement surgery and after the second year following hip replacement surgery. The third section of the thesis performs an extensive sensitivity analysis on the widely-quoted cost utility estimates for seven medical procedures, calculated by Gudex (1986). The estimates are shown to be sensitive to Gudex's conversion of health outcome data into the Rosser-Kind Classification, her assumptions concerning the survival period / life expectancy following each of the medical procedures and the selected discount rate. A more in depth analysis is then performed on the cost utility estimate for one of the seven procedures, ceftazidime treatment of cystic fibrosis. It is demonstrated that the health outcome and cost assumptions underlying the cost utility estimate for this procedure are not supported by the medical literature. Finally, the thesis raises a number of issues for discussion.