The main goal of this work to find correlation between the concentrations of radionuclides in outdoor air and the meteorological conditions like: air temperature, atmospheric pressure, wind velocity and amount of precipitation. Because the sampling period of radionuclides concentrations in air was relatively long (7 days), the average levels of meteorological parameters have been calculated within the same time. Data of radionuclide concentrations and meteorological data have been analyzed in order to find statistical correlation. The regression analysis and one of the AI methods, known as neural network, were applied. In general, analysis of the gathered data does not show any strong correlation between the meteorological conditions and the concentrations of radionuclides in air. A slightly stronger correlation we found for radionuclides with relatively short half-lives. The only positive correlation has been found between the Be-7 concentration and air temperature (at the significance level a=0.05). In our opinion, the lack of correlation was caused by a too long sampling time in measurements of radionuclides in outdoor air (a whole week). Results of the analysis received by means of the artificial neuron network are better. We were able to find certain groups of meteorological conditions, related with the corresponding concentrations of particular radionuclides in air. Preliminary measurements of radon progeny concentration support the thesis that the link between changes of meteorological parameters and concentrations of radionuclides in ambient air must exist.