Several time series forecasting models and artificial neural networks(ANNs) model are applied for Korean vegetable farm-level price forecasting. The application results show that time series forecasting models such as ARIMA and GARCH(GARCH(p,q), GARCH-M, TARCH, EGARCH, GARCH-t etc.) model outperform ANNs model in ex-post forecasting. Four criteria are applied to evaluate model performance: RMSPE(root mean square percentage error), MAPE(mean absolute percentage error), Theil;s inequality coefficient and TPFE(turining point forecasting error). Time series forecasting technics perform better than ANNs in terms of various out-of-sample forecasting. But ANNs model outperform time series forecasting technics in turning point forecasting error of onion price.