Abstract: Extreme value theory has been wildly applied in many fields, especially the multivariate extreme value distributions. Moment estimation is a classical estimation method because of its simple calculations. The paper considers the bivariate extreme value distribution in mixed model with exponential margins. The estimator and asymptotic variance of the dependence parameter are given. We also compare moment estimation with a maximum likelihood estimation in finite sample size. The results indicate that moment estimation is good for all practical purposes.