I have empirically examined the additional value relevance of two accounting risk measures to three traditional market risk measures, using current and future ROE and R regressions on those risk measures among the Chinese firms listed in Hong Kong, Shanghai and Shenzhen stock exchanges. My findings are as follows. Firstly, Hong Kong firms’ accounting risk measures, i.e., accounting beta(ABETA) and standard deviation of return on equity(ROESD) are remarkably higher than those of Shanghai and Shenzhen firms, while Hong Kong firms’ market risk measures including market beta(MBETA) are substantially lower than those of Shanghai and Shenzhen firms. Secondly, the Hong Kong firm’s adjusted R² in ROE and R regressions on both ABETA and ROESD is generally higher than that of Shanghai and Shenzhen firms. Especially, ABETA’s R regression coefficients of Hong Kong and Shenzhen firms are consistently and significantly positive, meaning ABETA as a risk measure works effectively on explaining current and future stock returns. Thirdly, ROE and R regressions on market risk measures do not show significant differences in the adjusted R² among firms listed in three Chinese stock exchanges. Fourthly, ROE regressions using both accounting and market risk measures together show that the adjusted R² of Hong Kong and Shenzhen firms is vividly higher than that of Shanghai firms. Finally, the mean difference analyses of adjusted R² between three factor regression models consisted only with market risk measures and five factor regression models, which additionally enter two accounting risk measures, show that Hong Kong firm’s adjusted R² is consistently and notably higher than that of Shanghai and Shenzhen firms over most test periods. My findings as a whole provide a new evidence on the empirical validity of the Chinese firm’s accounting risk measures by confirming their explanatory power on ROE and R variations.