Health-adjusted life expectancy (HALE) is one of the most attractive summary measures of population health. It provides balanced attention to fatal as well as non-fatal health outcomes, is sensitive to the severity of morbidity within the population, and can be readily compared between areas with very different population age structures. HALE, however, cannot be calculated at the small-area level using traditional life table methodology. Hence we propose a Bayesian random-effects modeling approach that recognizes correlations and pools strength between sexes, age-groups, geographical areas, and health outcomes. This approach allows for the calculation of HALE for areas as small as 2000 person years at risk and with relatively modest health state survey sample sizes. The feasibility of the Bayesian approach is illustrated in a real-life example, which also shows how differences in areas' health performances can be adequately quantified. Such information can be invaluable for the appropriate targetting and subsequent evaluation of urban regeneration, neighborhood renewal, and community-based initiatives aimed at improving health and reducing health inequalities.