It is accepted that within Britain there are persistent and growing in
equalities in mortality between groups of people as defined by their s
ocial class. This paper shows that similar persistent and growing ineq
ualities prevail between groups of people defined by district of resid
ence. Although there is some confusion between these two ways of group
ing people - there is a slight tendency for people of the same class t
o live in the same district. This paper reviews the geographical liter
ature which may shed light on why inequalities in mortality are wideni
ng between districts in Britain. We present new data for a set of 293
unchanging districts by amalgamating published reports from the 1950s,
60s and 70s with individual postcoded mortality records from the 1980
s and 90s, aggregated to the 293 districts using a Geographic Informat
ion System and Census data from 1971, 1981 and 1991 (including estimat
es of the residence of the 'missing million'). We find that a growing
proportion of premature deaths in Britain can be attributed to some as
pect of rising spatial inequalities. Changing geographical inequalitie
s in health are not simply a passive reflection of social inequalities
. To begin to investigate them, however, we first need to measure them
properly.