This study discusses the interaction between private information, overconfidence and intraday trading behaviour. Six thousand items consisting of 5-minute intervals of microstructure data in Taiwan market were collected and analysed during the period from September 2010 to the end of February 2011. Several important conclusions are obtained after Vector Autoregressive (VAR), Analysis of Variance (ANOVA) and Granger causality analyses of these data. First, when there is more private information in the market, the degree of overconfidence among investors is higher. Second, when private information and turnover rate are considered, the trading volume and return volatility show either a leading or a lagging relationship. Third, the appearance of private information is highest during opening and closing and has a mutual causal relationship with the trading volume. Fourth, the variation of private information reaches its maximum during opening, which is the primary factor that affects return volatility. Fifth, the January effect does not affect the sequence of trading volume and return volatility. Furthermore, return volatility is greatly influenced by the turnover rate; a high overconfidence among investors unfavourably impacts the stable development of the stock market.