Stockton-on-Tees has the highest geographical inequalities in health in England with the life expectancy at birth gap between the most and deprived neighbourhoods standing at over 17 years for men and 11 years for women. In this study, we provide the first detailed empirical examination of this geographical health divide by: estimating the gap in physical and general health (as measured by EQ. 5D, EQ. 5D-VAS and SF8PCS) between the most and least deprived areas; using a novel statistical technique to examining the causal role of compositional and contextual factors and their interaction; and doing so in a time of economic recession and austerity. Using a stratified random sampling technique, individual-level survey data was combined with secondary data sources and analysed using multi-level models with 95% confidence intervals obtained from nonparametric bootstrapping. The main findings indicate that there is a significant gap in health between the two areas, and that compositional level material factors, contextual factors and their interaction appear to be the major explanations of this gap. Contrary to the dominant policy discourse in this area, individual behavioural and psychosocial factors did not make a significant contribution towards explaining health inequalities in the study area. The findings are discussed in relation to geographical theories of health inequalities and the context of austerity.