We collected airborne particles of Japanese cedar pollen with Burkard's sampling tape in Toyama from February to April 1992. The tape was cut into two pieces in parallel to time axis. The one of piece of the tapes was stained with glycerin-jerry and stained pollens were counted with a microscope. The other piece was treated according to the immunoblotting technique. The airborne pollen allergens, reacting with anti-Cry j I monoclonal antibody, were stained as blue spots. The spots were classified by diameter into two groups, large spots (> 50 microns) and small spots (< 50 microns). There were significant correlations found between the airborne Cry j I allergen spots (in large and small) and actual pollen counts obtained with the Burkard's sampler and the Durham's sampler (r = 0.729, 0.586 in large spots and r = 0.676, 0.489 in small spots, p < 0.001). The counts of small spots stayed in high level even in April when actual pollen counts decreased. We concluded that this discrepancy was caused by allergenic crushed cedar pollen particles staying floating longer than actual pollens. Secondly we set a gauge of wind speed and direction at the same point as the samplers. The actual pollen counts and large spots counts were significantly larger in the wind (SE wind in Toyama city) from cedar trees blooming area than other areas. However small spots counts did not differ significantly according to wind directions. Wind speed did not effect on actual pollen counts, large spots counts and small spots count.