A previous study has shown a strong regional association between cold climate and coronary mortality in Sweden. The present study intended to elucidate further this association, in particular whether or not it could be attributed to smoking, hypertension and fat consumption. The 284 Swedish municipalities were used as units. The outcome was mortality (SMR) from acute myocardial infarction in men aged 40-64. Prevalence of smoking and use of snuff was estimated by interviews. Sales of antihypertensive drugs were used to estimate prevalence of hypertension. Sales of butter and estimates of consumption of saturated fat by interviews were used to estimate the consumption of fat. Weighted determination coefficients (D) were calculated. Apart from the strong association between cold climate and coronary mortality (D = 0.39), an association was shown between fat consumption and coronary mortality (D = 0.22). A weaker association was found between prevalency of snuffing and coronary mortality (D = 0.15) and between sales of butter and coronary mortality (D = 0.10) and the weakest between sales of antihypertensives and coronary mortality (D = 0.06). No association was found between prevalence of smoking and coronary mortality. When the other explanatory factors were added to cold climate in the model no substantial enhancement of the association was achieved. Cold exposure was correlated to the other explanatory factors. To conclude, this study showed that the strong association between cold exposure and coronary mortality was not affected by the regional variation in the estimates of fat consumption, hypertension or tobacco use.