Crop models are quantitative scheme for predicting the growth, development and yield of a given crop and these can be used as tools to forecast climatic scenarios and access management practices. Many crop models have been developed to help understand the process within cropping system and predict the crop yields as a function of weather and climate. Loss of water and nitrogen from light sandy loam soil of experimental site and climatic variability has been decreased the production of rice-wheat system, which provide the staple food to Indians. So, a field experiment merged with modelling study was conducted to find out the best management options for maximum and sustainable production of rice-wheat system. The treatments were combinations of rice and wheat crop residues and 6 application rates of fertilizer nitrogen. For the modelling study the DSSAT3.5 was adopted. The data related to weather, soil, crop management were archived. The models CERES-Rice and CERES-Wheat were calibrated and validated. Weather variables of 9 consecutive years were loaded and the weather generator program SIMMETEO was used to generate the future climatic scenarios. The monthly average of generated and actual weather data were compared, statistically tested and found well matched. So, the generated data was used to run Seasonal and Sequence analysis program of DSSAT in selecting the best treatment of rice and wheat for maximum production and stable productivity respectively. The predictions of Seasonal analysis revealed that DSSAT was dependable and capable model in selecting management options for maximum economic yields of both the crops. The results of Sequence analysis clarified good correlation in between the predicted precipitation and predicted crop yields. The incorporation of rice and wheat residues along with 100 and 80 kgN/ha for rice and wheat respectively was selected as the best treatment for sustainable production of rice-wheat system.