In order to forecast peacetime aviation ammunition consumption precisely, according to former consumption data, bench mark forecast was made. Multivariable linear regression forecast model, grey forecast model, shunt wound and series wound grey multivariable liner regression models were given. The result of each model had been tested. The series wound model can reduce the randomness of original data and improve forecasting precision;The shunt-wound model can integrate many factors and has a useful value. The series wound model has the most precise forecast value; it can be used for practical forecast and aviation ammunition peacetime indent.