Autopsy cases (n = 358) were divided into hypothermia, suspected hypothermia, and control groups. The catecholamine levels did not correlate with the length of the postmortem period. The adrenaline to noradrenaline ratio was most effective in detecting hypothermia (68.9 percent sensitivity, 78.1 percent specificity). The median adrenaline concentrations were significantly higher in hypothermia than in control groups. The control group containing mostly sudden cardiac deaths with no cold exposure had a noradrenaline level comparable to the hypothermia groups. The sensitivity and specificity of determining Wischnewsky's spots in hypothermia deaths were 63.9 percent and 88.3 percent, respectively. The adrenaline to noradrenaline ratio is more suitable in proving antemortem cold stress than either of these independently, and its diagnostic value is comparable to that of Wischnewsky's spots. (Published Abstract)