This paper examines the herd behavior in six segmented markets on the Chinese stock markets. Using the OLS, GARCH, Quantile Regression, and State Space Models to examine the daily returns from 2003 to 2018, we find that herd behavior exists widely in all the segmented markets examined in China, particularly in the two B-share markets. The two B-share markets also show stronger (weaker) asymmetric herd effects when market returns are rising (falling) and when trading volumes are higher (lower). Further evidence suggests that the herd effect in China became stronger during the period of the Chinese stock market turbulence. The results can provide enlightenment for the Chinese policymakers to stabilize and to improve the efficiency of stock markets, and also help investors to identify their markets of interest and control financial risks.