Rice, being the staple food grain of Odisha, holds a crucial place in the state’s economy and food security. Rice holds around 69% of the total cultivable area in Odisha, making it crucial to have an accurate forecast of its status for stakeholders in agriculture. Modelling and forecasting of time series dataset of yield and production of rice from 1970-71 to 2019-20 is carried out in this study, using Auto Regressive integrated Moving Average (ARIMA), Artificial Neural Network (ANN) and Hybrid ARIMA-ANN methodologies. ARIMA is a linear modelling approach where whereas ANN is more of a non-linear modelling technique. The hybrid ARIMA-ANN methodology integrates the strengths of both models to effectively capture both linear and non-linear patterns within the dataset under study. It was found that ARIMA(1,1,1) with constant and under the developed ANN models, the Neural Network Autoregression(NNAR) of order NNAR(3,2) came out to be the best fitted model for both of the variables under study. ARIMA(1,1,1)-NNAR(1,1) is found to be suitable for both yield and production of rice in Odisha. All three models are compared using accuracy measures like RMSE and MAPE, and the hybrid methodology is found to be superior to others.