Rice blast is a detrimental disease widely prevalent in the world. Many disease predictive models have been developed for Rice blast disease (RBD) in different regions, but the disease prediction modeling in semi-arid areas is less characterized. Here, we focused on developing a disease predictive model (DPM) for rice blast disease (RBD) based on eight-year (2009–2016) environmental variables and their correlation with the RBD severity data, followed by validation with two-year (2017–2018) field trial data in a semi-arid region of Pakistan. Besides the correlation analysis, the model implied multiple linear regression analysis. The DPM indicated that five environmental variables maximum (Max) and minimum (Min) temperature, rainfall (RF), relative humidity (RH) and windspeed (WS), significantly contributed to the development of RBD during eight years. The obtained values of standard statistical indicators, coefficient of determination (R2) low standard error (SE) ≤ 10, and significance of F-distribution of regression statistics proved the effectiveness in predicting RBD for eight years. Furthermore, the criteria of root-mean square error (RMSE) and error (%) of 8 year data and observed data suggested a striking closeness between old and observed values of RBD severity, indicating the effective reliability of the model in both contexts. The analysis of 8 year DPM data showed that five environmental variables (max and min temps, RF, RH and WS) could cause up to 92% variability in RBD. Except for min temp, which showed a negative correlation suggesting a minor influence on disease development, the other four environmental factors exhibited a positive correlation with RBD severity, indicating the maximum contributing factors in disease development. The developed model helped us to predict the ranges of environmental factors, Max and Min temp 40–42°C and 22–24°C, RF 2-2.5 mm, RH 50–70%, and WS 9–11 Km/h, significantly favorable for RBD. The current prototype DPM of RBD has the potential for practical application if combined with a weather prediction system, which could be useful in developing rice blast disease warning system in the semi-arid zone of Punjab and predicting the appropriate time of the fungicidal sprays.