連續時間短期利率模型在衍生性商品訂價與風險管理上至為重要，本文估計並比較連續短期利率模型在台灣短期利率上之實證表現。本文主要貢獻在於使用不同的計量技巧來比較短期利率模型的實證表現。在參數估計方面，本文使用兩種近似方式將連續的利率過程寫成間斷模型，再以一般動差法(GMM)與準最大概似法(QML)估計，以評估模型在資料配適上的優劣；此外，亦比較不同資料頻率（月、週資料）對模型估計與評估是否有顯著的不同。同時，為說明估計結果之穩健性，本文亦針對美、日二國短期利率作探討。主要實證結果指出，不同近似方法所得到的估計結果並無太大差異，不過估計方法與資料頻率會影響估計結果，使用週資料並以準最大概似法估計所得到之結果較具穩定性與效率性；其次，台灣短期利率存在均數復歸現象，且短期利率波動性受利率水準值影響，惟敏感度係數估計值小於1，而在模型評估方面，相較於一般化的未受限制模型，以CIR-SR模型在資料的配適上表現最好；美日兩國實證結果亦指出估計結果受估計方法與資料頻率之影響，而美國短期利率並不存在均數復歸現象，日本短期利率則僅在以QML估計時始具有均數復歸現象，兩國短期利率之波動性亦均受利率水準值影響，只是敏感程度不同，模型的配適則均以CEV模型表現最佳。

We utilize GMM and QML to estimate and compare the performance of continuous-time short interest rate models for 30-days Taiwan Commercial Paper rate (CP2) and one-month Japan and American interbank interest rates. We discretize the continuous-time models by using two different approaches, and then use weekly and monthly data to estimate the parameters. The models are evaluated by data fit. We find that the estimated parameters are similar for different discretization approaches and would be more stable and efficient under QML with weekly data. There exists mean reversion for Taiwan CP rate but not for Japan and American interbank rates. The relationship between the volatility and the level of interest rates are less than 1 for all the three countries and smaller than that of American T-Bill rates reported by CKLS (1992) and Nowman (1997). We also find that CIR-SR model performs best for Taiwan CP rate and CEV model performs best for Japan and American interbank interest rates.