The present study critically examined the influence of weather parameters on the initiation and spread of the cucumber (Cucumis sativus) downy mildew disease and developed a suitable weather based disease forewarning models. The field experiment was laid out in the farm of Bidhan Chandra Krishi Viswavidyalaya, Kalyani, West Bengal during 2008-2011. Cucumber crops were sown during 33 times throughout the entire span of four years covering all the growing seasons of the years. High humidity (RH>94%) and average temperature (24-300C) along with leaf wetness not less than 8 hrs was found to trigger the initiation of downy mildew disease of cucumber. A model was produced using logistic regression analysis and cumulative value of night leaf wetness duration, average night temperature, night relative humidity (RH) and number of night hours having RH>95% from sowing time were a significant disease predictor. Among the weather parameters daily mean temperature during growth period had maximum degree of association (r = -0.50) with percent disease intensity (PDI). Disease progress curves were presented using logistic and Gompertz model. W eather based prediction model has been developed with different weather indices and disease severity using weekly average value and cumulative value from date of sowing. Cumulative values of weather variables could explain 95% variance of disease intensity, whereas average values of weather variable could explain only 33 % variance of disease intensity. These results will improve the timing and application of the fungicide spray for the control of cucurbit downy mildew.