Abstract: Stock market network was constructed by using the minimum spanning tree method to study the evolution of MST structure， the topological indicators and individual behavior characteristics of the key nodes at different periods before， during and after the extreme volatility of stock indexes. The results showed that the China′s stock market industry has a strong concentration and industry-wide homogeneity. The MST network shrinks during extreme volatility， and the network expands after volatility. In the middle and late periods of extreme volatility， the sudden emergence of the materials industry and information technology industry occupies an important position in the network. The extreme volatility caused by different infectious sources has different ways of transmission risk. Extreme volatility in stock indexes exacerbates the linkage between stocks， assets are reconfigured in the market and the network structure is undergone new adjustments.

Key words: minimum spanning tree method, stock index extreme volatility, stock market, MST network, topological structure