The mortality rates of the Moscow population in the summer months of 2007-2011 were analyzed. The most significant unfavorable impact of abnormally elevated temperature of air and its pollution on mortality rates in the Moscow population was recorded in the summer months of 2010, causing a considerable rise in mortality rates (11 541 more deaths), particularly in the age group of 70 years or older. The proportion of this population group in the age structure of mortality rose from 58.7% (in the January-May 2010 period) to 63.7% (in the summer of 2010) (p<0.0001). The negative weather conditions led to a dramatic (1.3-2.5-fold) increase in death rates among people with chronic diseases of circulatory, respiratory, nervous, and urogenital systems in the July and August of 2010 versus those in the same months of 2002-2009 and 2011. The predictive power of 3 linear regression analysis models for mortality rates was compared, by taking into account basic interactions (year, season, month, and day of week), temperature of air and its pollution indicator PM10. The highest predictive power (multiple correlation coefficient R2=0.818) was proven for a model considering the threshold temperature of 20 °C and its excess per day of forecast and previous 5 days (lag 5 days) that taking into account the air PM10 concentration. That for a model considering the threshold temperature of 25 °C and its excess per day of forecast (without any lag) was R2=0.442. The predictive power for a model accounting the threshold temperature of 25 °C and its excess per day of forecast, without any lag, with consideration for PM10 concentration, was R2=0.594.