Forecasting the time when the atmospheric pollen season of allergenic plants begins is particularly important for doctors and their patients. The aim of this paper is to determine whether it is possible to forecast the start of the oak (Quercus) pollen season in Rzeszów, Poland. In the elaboration of the most effective model, various forecasting techniques were tested: growth degree days (GDD degrees C); meteorological factors; bioclimatic factors; and indicator taxon. The aerobiological monitoring was carried out in 1997-2005 and 2007 in Rzeszów (SE Poland). In the presented investigation, three methods defining the start of the Quercus pollen season were selected on the basis of accumulated sums of pollen or the constant occurrence of pollen grains in air. Despite the application of different combinations of GDD degrees C methods and threshold temperatures, the correlation coefficients between the expected and obtained values were low. In some cases, however, they proved highly effective for the test years (2005, 2007) with the accuracy of a few days. For GDD degrees C methods, the best threshold temperatures range between 5 and 6 degrees C. Models based on bioclimatic indices and meteorological variables were not satisfactory. On the basic of the 10 years of results, the method of indicator species were good for forecast the start of oak pollen season. Birch was the best indicator taxa.