This paper tests the efficiency of the KOSPI 200 stock index options market by examining the possibility of abnormal returns from arbitrage trading strategies using volatility forecasts. The performance from an arbitrage trading using volatility forecasts depends on the accurate volatility forecasting and the accurate option pricing model. In this paper, two forecasting methods are used for the test; one is GARCH forecast method based on historical return data and the other is Implied Volatility Regression(IVR) forecast method based on implied volatility data. It is assumed that the volatility which an arbitrage trading uses is not constant. This paper shows the different results of the arbitrage trading when Black and Scholes pricing model with constant volatility is used and when Hull and White pricing model with stochastic volatility is used. The trading strategy in this paper implies that the trader buys(sells) options that are underpriced(overpriced) and delta-hedges the position by selling(buying) index portfolio. We find that IVR predicts more accurately than GARCH(1,1) does and that the combination of IVR and Hull and White pricing model leads to potential earnings that significantly exceed transaction costs. In particular, from out of the money options trading, the trader can earn an abnormal profit exceeding the assumed transaction cost suggesting that KOSPI 200 stock index options market is not efficient.