A mathematical model for point source emissions, which allows to determine the ecological risk, improve the ecological situation in the region with the ability to manage risk. Environmental risk is defined as the probability of exceeding a multidimensional integral field concentrations of pollutants its Maximum Admissible Concentration (with a small number of measurements) or frequency exceeding the given measured concentrations of pollutants its Maximum Admissible Concentration (with a large number of measurements). Components of air pollution regions, cities, enterprises are individual emission sources, so primarily developing assessment and environmental risk analysis should be performed for individual sources and, above all, to prevent significant risks – for the prognostic assessment of newly constructed or reconstructed enterprises, and for existing facilities – from the measurement data. For a single point source mathematical model includes: – raw data (results undertorch measurement characteristics of pollutants, sanitary protection and residential areas, the design parameters of the source and characteristics of the external environment for the worst-case); – forecast concentrations of pollutants depending on the design parameters of the source and characteristics of an environment for a source with circular and rectangular mouth, hot and cold emissions, extremely dangerous low wind speeds; – depending on the definition of the amendments to the measured concentrations due to the difference of the worst conditions from measurements; – depending on the definition given to the worst conditions of the measured concentrations and their statistical processing in order to obtain: a) the numerical characteristics of the density distribution of the concentration of pollutants emitted (mathematical expectations, standard deviations and correlation coefficients); b) environmental risk α, defined at a relatively small number of trials as a multidimensional probability integral of the density distribution of the concentrations obtained with the numerical characteristics, and a large number of tests - the frequency of the measured concentrations exceeding; c) environmental risks from air pollution αj separate pollutants; d) depending on the definition of the numerical characteristics of the forecast distribution density is not measured (secondary) concentrations of pollutants emitted to address their risks when determining α and αj. It is shown that, in accordance with applicable regulatory requirements must consider the risk of the negative impact of air pollution to humans, the level of which is determined using the values of the maximum one-time maximum allowable concentrations pollutants. Defined by experienced stochastic patterns of distribution of contaminants in the atmosphere. Studied the random variation of concentrations. Theoretically and experimentally substantiated principles of developing a mathematical model to assess the actual risk from ecological point sources of emissions.