This study attempts to forecast production and yield of two main cash crops namely sugarcane and cotton crops of Pakistan by using Auto Regressive Moving Average (ARMA) and Auto Regressive Integrated Moving Average (ARIMA) models of forecasting. Using data for 1948 to 2012, productions and yields of both crops were forecasted for 18 years starting from 2013 to 2030.ARMA (1, 4), ARMA (1, 1) and ARMA (0, 1) were found appropriate for sugarcane production, sugarcane yield, and cotton production respectively, whereas ARIMA (2, 1, 1) was the suitable model for forecasting cotton yield. Some diagnostic tests were also performed on fitted models and were found well fitted.