Effect of weather factors on fluctuations of spore population of Pyricularia grisea and the occurrence of the disease was considered. During growing seasons of 2006–2007, paddy fields were chosen in distance of five kilometers from weather stations of Rasht, Lahijan and Anzali in Guilan province and spore population (Ps) were measured daily using sporetraps. Weather data including precipitation (P), daily maximum and minimum temperature (Tmax, T min), daily maximum and minimum relative humidity (RHmax, RH min) and sunny hours (SH) were obtained from weather stations. The relationship between spore population fluctuations and weather data was analyzed and the most important weather factors affecting spore population and predicting blast were determined. Accordingly, weather factors such as P, T max, RH min and SH are the most important factors predicting rice blast in Guilan and enough precipitation, increased daily RH min. , decreased daily T max and SH result in increased spore population and blast occurrence during next 7–10 days. To predict final leaf blast severity (Y flbs) and neck blast index (Y nbi ), factors such as T max, T min , T, RH max , RH min, RH , P and SH and Ps were used for modeling. For leaf blast, these factors were considered for June and July and for neck blast, the same factors used for August. Step wise regression was applied for modeling. Statistics like r, R 2 , aR 2, SE, F and Durbin-Watson were applied for evaluating the models. Finally, the two quantitative models: flbs = –2.41–2.80 Tmin +0.68RHmin–0.015Ps–0.014P+0.052SH (R 2 = 96.73%) and Y nbi = –24.11+0.08T max +0.19 RH max +0.034Ps–0.015P+0.016SH (R 2 = 73.97%), were introduced for predicting final leaf blast severity and neck blast index, respectively. Related to effects of amount of applied N fertilizer (F) and date (D) and space (S) of transplanting, the results showed high correlation between F and Y flbs and Y nbi, but such high correlation was not observed for D and S. The best function for predicting Y flbs was Y = 4.46–4.12F+1.93F 2 (R 2 = 96.37).The best equation for predicting Y nbi acquired when F, D and S were applied in multiple regression, Y = 2.06+0.33F+0.10D–0.03S(R 2 = 54.40)