The data on concentration of some pollutants (CO , SO2 , NO2 , dust, phenol) measured in the city of St. Petersburg by means of the mobile ecological laboratory have been used for construction of the distribution function of these substances. The latter has been used for estimating the probability of exceeding the maximum permissible concentration (MPC) of these substances. The important property of the distribution function is revealed. It is the same (automodel) within the limits of measurement error and the accuracy of calculation (taking into account the limited number of samples) either for different ingredients or for different observation sites, if it is constructed for the deviation of the concentration from the average value normalized to the root–mean–square (standard) deviation. Most exactly the empirical distribution function can be approximated by the lognormal distribution.