Two particle samplers for ambient air, situated together: a static size-selective bio-aerosol sampler (SSBAS) and a Burkard pollen and spore trap were compared in sampling intact birch pollen grains through one flowering period of Betula (a total of 44 days). The SSBAS trapped pollen grains three times more efficiently than the Burkard trap, but the variations in pollen counts were significantly correlated. In contrast, birch pollen antigenic activity and the pollen count in the Burkard samples were not closely correlated. The antigenic concentration was occasionally high both before and after the pollination period. There was a high birch pollen antigenic activity in particle size classes where intact pollen grains were absent, even on days when the pollen count was very low. Correspondingly, on days with high birch pollen counts in the air, pollen antigenic activity was on several occasions low, indicating that pollen grains were empty of antigenic material. The small particle size classes are especially important to allergic patients because they are able to penetrate immediately into the alveoli and provoke asthmatic reactions. Therefore, aerobiological information systems based on pollen and spore counts should be supplemented with information concerning antigenic activities in the air.