We develop the bullwhip effect measure under a bivariate VAR(1) demand model. The formula can simultaneously explain the interaction of all parameters concerned. The correlation coefficient of the two demand’s errors affects the bullwhip effect. The conditions at which the bullwhip effect might not exist, are developed.