Scotland, which has one of the highest death rates in the world from coronary heart disease, also has considerable regional variation in mortality from this cause. The relation between standardised mortality ratios for coronary deaths (1979-83) for 56 local government districts and a range of socioeconomic factors from the 1981 Census as well as climatic factors and water hardness were investigated. Strong associations were seen with several measures of social disadvantage, the strongest being with percentage of male unemployment. A fitted multiple regression model with mortality from coronary heart disease in men found independent effects of two social variables (percentage male unemployment and percentage social class III-V) and one climatic factor (rainfall). The model explained much (73%) of the geographical variation in mortality from coronary heart disease, but part of the geographical pattern, in particular some of the east-west gradient in mortality, remained unexplained by it. Explanations for the geographical variation and the association with economic factors are currently being sought in terms of individual risk factors in a large screening study, the Scottish Heart Health Study.