The present study is an attempt to forecast the total export (at constant 2010 US $) of India for twenty years by applying relevant univariate time series ARIMA model. The data was collected from 1960 to 2016 using World Bank National Accounts data, and OECD National Accounts data files. Augmented Dickey Fuller (ADF) test and Phillip-Perron (PP) test has been used to test stationarity of the series. Several possible ARIMA models has been run with reference to the fitted autocorrelation function (ACF) and partial autocorrelation function (PACF) and relevant model, that is in this case ARIMA (0, 1, 3) has been selected using minimum Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) and forecast has been carried out together with upper and lower 95, 90, 80 percent confidence interval from year 2017 to 2036. The residual ACF and PACF have been used for diagnostic check of the forecast. The forecast reveals that the future export will be rising in next twenty years.