Using principal component analysis, this paper separately constructs A-share's and H-share's composite indexes as proxy for market investor sentiment. In the two-pass conditional asset-pricing framework, this paper allows the Beta varying with investor sentiment and firm characteristics. The empirical results indicate:after incorporating investor sentiments in conditional pricing models, the size effects of A-share market and H-share market are no longer significant; meanwhile, the B/M effect of A-share market becomes less significant. Therefore, this paper draws the conclusion that incorporating investor sentiment as conditioning information in conditional asset-pricing models improves the model performance in capturing asset-pricing anomalies, however, the improvement on A-share pricing is stronger than that on H-shares, which demonstrates that different market investor sentiments have different degrees of impacts on asset pricing, resulting in the price difference of dual-listed A and H-shares.