Weather is an instantaneous condition of atmosphere over a place. Weather is highly variable and constantly changing from day-day or hour-hour. Due to global warming and human interference in nature has made the predication of climate, a totally unpredictable. Forecasting of weather has been a very important issue through last two decades. Soft computing techniques are being increasingly used for forecasting of various systems. Soft computing models composed of fuzzy logic, neural network, evolutionary computing, genetic algorithm etc., Here Fuzzy logic and neural network has been selected as the soft computing technique. This paper presents a comparative study of a traditional statistical time-series model and a neuro-fuzzy model. The performance of the selected models are compared and evaluated using the statistical estimators such as Moving Average Error (MAE), Root mean square error (RMSE) and Coefficient of determination (R2). The experimental results indicate that neuro-fuzzy model provides a good precision in prediction in terms of statistical indicators. © 2015 Elsevier B.V., All rights reserved.