One of the potential effects of an anthropogenically induced climate change is a change in mortality related to thermal stress. In this paper, existing literature on the relationship between average temperatures and mortality is evaluated. By means of a simple meta-analysis an aggregated effect of a change in temperature on mortality is estimated for total, cardiovascular and respiratory mortality. These effect estimates are combined with projections of changes in baseline climate conditions of 20 cities, according to climate change scenarios of three General Circulation Models (GCMs). The results indicate that for most of the cities included, global climate change is likely to lead to a reduction in mortality rates due to decreasing winter mortality. This effect is most pronounced for cardiovascular mortality in elderly people in cities which experience temperate or cold climates at present. The sensitivity of the results to physiological and socio-economical adaptation is examined. However, more research is necessary to extend this work by inclusion of data from a wider range of populations.