OBJECTIVES: There is increasing awareness that equity weights should be used to recalculate the value of QALY gains for different patients. It is unclear however how these equity weights should be determined: on the basis of health prospects (for instance "rule of rescue"), fair innings, or on a combination of both, for instance an equity concept that has been referred to as proportional shortfall. To answer this question, we compared the observed rank order of 10 conditions with the theoretical rank orders that were predicted by each equity concept. METHODS: 60 respondents (students, researchers, health policy makers) rank ordered 10 conditions using the paired comparison technique. This observed rank order was compared to the rank orders expected on the basis of the equity concepts fair innings, prospective health and proportional shortfall. To allow for comparison of the conditions in terms of each equity concept, we described the conditions in terms of age, disease free period, duration of disease, quality of life, and life years lost. RESULTS: The observed rank order of the 10 conditions was best predicted by the fair innings concept (corr.=0.908, p<0.01). Proportional shortfall was also well correlated with the observed rank order of the conditions (corr.=0.780, p<0.01), but prospective health was not statistically significantly related. This is remarkable, as it has often been suggested that the "rule of rescue" is the most important determinant of the distribution of health care. CONCLUSIONS: Measurable interpretations of equity make it possible to test the importance of concepts of equity in the allocation of health care. The fair innings argument and proportional shortfall may provide a basis for determining equity weights for recalculating the value of QALY gains for different patients. When put to a critical test, the prospective health argument is out weighted by arguments resembling fair innings.