We examined the effect of the wind vector analyzed into its three components (direction, speed and persistence), on the circulation of pollen from different plant taxa prominent in the Thessaloniki area for a 4-year period (1996-1999). These plant taxa were Ambrosia spp., Artemisia spp., Chenopodiaceae, Corylus spp., Cupressaceae, Olea europaea, Pinaceae, Platanus spp., Poaceae, Populus spp., Quercus spp., and Urticaceae. Airborne pollen of Cupressaceae, Urticaceae, Quercus spp. and O. europaea make up approximately 70% of the total average annual pollen counts. The set of data that we worked with represented days without precipitation and time intervals during which winds blew from the same direction for at least 4 consecutive hours. We did this in order to study the effect of the different wind components independently of precipitation, and to avoid secondary effects produced by pollen resuspension phenomena. Factorial regression analysis among the summed bi-hourly pollen counts for each taxon and the values of wind speed and persistence per wind direction gave significant results in 22 cases (combinations of plant taxa and wind directions). The pollen concentrations of all taxa correlated significantly with at least one of the three wind components. In seven out of the 22 taxon-wind direction combinations, the pollen counts correlated positively with wind persistence, whereas this was the case for only two of the taxon-wind speed combinations. In seven cases, pollen counts correlated with the interaction effect of wind speed and persistence. This shows the importance of wind persistence in pollen transport, particularly when weak winds prevail for a considerable part of the year, as is the case for Thessaloniki. Medium/long-distance pollen transport was evidenced for Olea (NW, SW directions), Corylus (NW, SW), Poaceae (SW) and Populus (NW).