In domain adaptation, entropy minimization are widely used. However, entropy minimization will bring negativetransfer when the pseudo-labels are inconsistent with the real labels. We hope to increase pseudo-label accuracy to counternegative transfer in entropy minimization. To this end, we introduce domain adversarial training into entropy minimization.Furthermore, we consider the misalignment caused by domain adversarial training under severe label shift. Therefore, wepropose method called entropy minimization and domain adversarial training guided by label distribution similarity. Throughdomain adversarial training which focus more on class-aligned divergence, our method improves pseudo-label accuracy andreduce negative transfer in entropy minimization. Extensive experiments demonstrate the effectiveness and robustness of ourproposed method.