The Kingdom of Saudi Arabia is playing a key role in the production of petroleum products. Accurate forecasting of the demand/ supply of these products is very essential for proper planning of the refineries of the Kingdom. An overview of the various forecasting methods used for this purpose is presented and a relatively new time series method, dynamic data systems, is applied. Autoregressive moving average models are developed and used for forecasting. A number of statistical tests are applied lo demonstrate that these forecasts are highly accurate and dependable