Using the data collected for the year 2004 in the Belgian Health Survey, we aim to show that, controlling for all the determinants of self-reported health state, there is a residual effect of geographic location on the self-assessed health, namely a more favourable scoring for individuals that are resident both in the region of Brussels and in the region of Walloonia, with respect to individuals that are resident in the Flanders. Regional effects do not change either if we take account of supply of health services or if we control for their utilization. Moreover the effect of past level of health is encompassed to test for habituation, and the results still hold. The above findings can be used both to construct “equivalent expected QALY’s or EEQ” ,i.e. the average quality adjusted life years that a newborn, taking account of the different average level of health at the regional level, can expect to experience in the different geographical areas of Belgium (the three regions have approximately the same EEQ), and to predict a “need factor” to be used either for equity analysis or to ascertain its evolution in time. ; info:eu-repo/semantics/published