In econometric time series analysis data which have high volatility would be very risky to be used as a basis for forecasting including the volatility of food prices in Indonesia. Time series data have a tendency to have a constant confounding error variance over time. Appropriate econometric model to estimate such behavior is called the Autoregressive Conditional Heteroscedasticity (ARCH) model and the Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model. This paper attempts to use ARCH/GARCH models to explain the behavior of food price inflation in Indonesia in time period of 2005.1 to 2010.6. It is explained that by incorporating elements of ARCH/GARCH better estimates will be achieved.