The relation between temperature and violence was found in many studies. However, the results of such studies demonstrated only that uncomfortably hot temperatures increase violence. There seem to be no data on the effect of cold temperatures. We studied the relation between temperature and violence for the Russian Federation because the Russian Federation is a country with huge climatic differences. Two types of the analysis of the data were applied. In Analysis 1 average yearly temperatures were used. For violent crimes a decrease in temperature resulted in the increase of the crimes after taking into account three socioeconomic variables. Analysis 2 was based on monthly data. Violence was high in winter and spring months but low in autumn months. In our opinion, the conventional models that are used to clarify the effect of hot temperatures cannot explain our results. We hypothesize that long periods of cold temperatures can be considered as mild chronic stress. Chronic stress may exert depression and depression is associated with irritability and anger. In some situations these emotions may stimulate violence. An increase in violence associated with city living and economic downturns may partially be a consequence of mild chronic stress.