Forecasting the dates on which maximum airborne pollen concentration occurs is very important for human health because of the well-known allergenicity of pollen. Airborne pollen concentration varies in response to the flowering rhythms of plants and to the meteorological factors. Meteorological trend affects the timing occurrence of the flowering phase, the concentration and the dispersal of pollen in the air. One of the main meteorological factors affecting flowering in Mediterranean plants is temperature. In this work several heat summation methods based on air temperature were used, in order to predict the time of higher pollen concentration of Olea europea L. in the air. Data of pollen airborne concentration were recorded for 18 years in a urban area of North Sardinia (Italy) with a Burkard seven-day recording volumetric spore trap (1984-2001). During the same period, air temperature values were also recorded by an automatic weather station located near the spore trap. Degree days were calculated for 15 years (1984-1998) using the daily averaging method and the single triangle method from nine starting dates (December 1st and 15th , January, 1st and 15th , February 1st and 15th March 1st and 15th , and April 1st), and five temperature thresholds (0°C, 4°C, 8°C, 12°C, and 14°C). The standard deviation in degree days, the coefficient of variation and the standard deviation in days were also calculated in order to determine the starting date and the threshold value that provide the best estimate of the dates. The analysis carried out showed that the single triangle method best predicts the peak of pollen concentration date with a difference of 3.13 days between predicted and observed dates using 0 °C as base temperature. In addiction, the 1999–2001 temperature and airborne pollen concentration data were used to test the accuracy of the methodology in predicting peak dates of pollen concentration. The method provides a good prediction with small differences between forecast and actual dates.