In this paper, we investigate the relationship among trading volume, volatility and jump based on Hu-Shen300 index high frequency data.We measure volatility using realized volatility proposed by Andersen and Bollerslev (1998) and decompose it into upside and downside parts by the asymptotic properties of realized upside and downside power variation defined in Bi, Zhang and Wu(2013).To measure jump component, we use a novel jump measure called signed jump proposed by Patton and Sheppard(2011).Our empirical analysis show that trading volume are positively related with realized volatility,RUPV and RDPV.We also find that trading volume is negatively related with signed jump.