Abstract Using daily data from July 1993 to June 2013, we estimate a VAR-BEKK model and find evidence of return and volatility spillovers among the different stock markets in the Greater China region (China, Hong Kong, and Taiwan). We apply the volatility impulse response function proposed by Hafner and Herwartz (2006) to quantify the impact of financial crises on expected conditional volatility. We observe that financial crises have positive and large impacts on expected conditional variances, but the size and dynamics of the impact is largely market specific. We specifically find evidence in favour of increased market integration. Consequently, had a financial crisis occurred recently, the impact of it on expected conditional variance would have been significantly lower nowadays compared with the initial dates when the financial crisis occurred. Finally, we estimate the density of the VIRF at different forecast horizons. These fitted distributions are asymmetric and show that large increase in expected conditional volatilities are possible even if their probability is low. These features denote good efficiency and reaction to shocks from market participants.