A time-varying multiple Copula model was constructed, and the Monte Carlo simulation technology was used to calculate VaR (Value at Risk) in order to accurately measure the risk resulting from the four exchange rates: CNY/USD, CNY/EUR, CNY/JPY and CNY/HKD of commercial Banks. Then, a comparative analysis was made on the measurement effect between the static multiple Copula-VaR and the time-varying multiple Copula-VaR. The results show that the connections between these exchange rates are indeed time-varying related, and the measurement effect of the time-varying multiple Copula-VaR is much better.