Central banks conduct monetary policy on a monthly basis and have interests in the dynamic analysis of monthly frequency data. However, GDP is observed quarterly and thus an impulse response analysis using quarterly model is conventional. A realistic approach is to estimate a monthly VAR using a mixed frequency data. In this paper, following Kim and Park(2007), we conduct a dynamic analysis of the effect of Korean monetary policy on GDP under VAR model with a mixed frequency data. According to the impulse response analysis based on GMM estimation, we find the increase of money supply affects GDP growth and the effect disappears gradually. However, the inflation and Won/Dollar exchange rate are not much affected by the money supply shocks. This approach may be useful especially during financial crisis because the monetary policy effect is analyzed by monthly period.