The present study was conducted to know the statistical investigation of price behaviour of potato in Agra. Potato is one of the most important food crops of India. The information on price of potato in Agra was collected for the period January 2004 to December 2012. Potato wholesale prices of Agra market were analysed using univariate ARIMA (2,1,1) model. Seasonal indices calculated showed that generally the price is low from November to January and it starts picking up from February, and reaches the maximum in November. Based on the Shwartz Bayes Criterion (SBC) and Akaike Information Criterion (AIC), the estimated best model was ARIMA (2,1,1). Short term forecasts based on this model were close to the observed values. The accuracy measures like MAPE and MSE are considered as the best models to forecast monthly prices. Among all the models, ARIMA (2,1,1) model was best with least Mean Average Percentage Error (MAPE) and Mean Square Error (MSE) value for Agra 20.57, 102.22 respectively. The data analysis is carried out using SAS 9.3

Keyword: Price behaviour; ARIMA Model; Forecasting, Seasonality; Stationarity; SAS; Wholesale price;