The instability of the producers and investors income coming from price fluctuation is a problem whose characteristics and causes should be thoroughly investigated given the importance of the commodity in the national agribusiness and their losses in terms of profitability, jobs and exchange value to Brazil. Given that, it is used the class of autoregressive conditional heteroscedasticity models (ARCH and GARCH) to characterize and analyze the volatility of the time series of monthly returns of soy, coffee, corn and fat ox. This analysis shows that these products are marked by having high price fluctuations, in which positive or negative shocks generate impacts in the long run. The sum of the volatility reaction and persistence coefficients showed values close or larger than one, indicating that the shocks up on volatility will last for a long period of time.