In order to contribute to prevent Japanese cypress pollen allergies, we investigated the relationship between the visually-observed amount of staminate flowers and the amount of dispersed pollen in cypress forests in Kanagawa Prefecture. The index of staminate flower quantity based on visual inspection in a cypress forest, showed a significant correlation with the number of staminate flowers collected in staminate flower traps. When comparing the index of staminate flower quantity with the number of pollen grains dispersed in Kanagawa and Tokyo, exponential regression revealed a significantly high correlation of 0.94 at Kanazawa Ward in Yokohama and 0.88 at Chiyoda Ward in Tokyo, demonstrating that pollen dispersion can be predicted. When we investigated the correlation between summer meteorological factors (hours of sunshine, average temperature, and precipitation) and cypress pollen dispersal, the only significant factor was precipitation from June to July, with the correlation of about 0.6, which was lower than the correlation with the index of staminate flower quantity. This demonstrated that it is difficult to predict pollen dispersal based on meteorological factors only. The correlation between the amount of cedar pollen and that of cypress pollen released over the previous 12 years was not significant, indicating that the pollen release rates of cedar and cypress are not synchronized. Our findings indicates that forecasting cypress pollen dispersion requires only visual inspection of cypress trees, and that forecasting based on meteorological factors is difficult, and forecasting based on visual observation of the index of staminate flower quantity is effective.