臺灣自2000年通過農業開發修正條例後，正式開啟農地自由買賣的紀元，結果對農村地景產生衝擊。為了探究農村地景變遷情形，並預測未來建地之空間分布，本研究以宜蘭縣三星鄉尾塹村、大洲村、大義村為研究區域，以馬可夫模式預測未來地景變遷，並選擇6項影響農村建地分布的因子，以邏輯斯迴歸模式進行建地之空間分析，最後整合馬可夫模式與邏輯斯迴歸模式預測未來建地空間分布情形。研究結果有幾項發現：（1）2000－2009年農地轉變成建地的速度較1987－2000年快，顯示農業政策對農村地景確有影響；（2）若以2000－2009年開發速度持續開發本區，至2054年將由建地取代農地成為本區主要土地利用方式；（3）距高速公路以及區內主要道路愈近之土地，愈容易變遷為建地，顯示交通為影響本區建地分布之主要因子；（4）由2018年、2036年以及2054年之預測圖發現，建地在空間分布有擴增的趨勢。農地急速消失所帶來的影響，除了未來可能出現農業生產匱乏的隱憂之外，農村生態系統的維繫、開放空間的供予以及優美景緻等價值，亦將隨著農村地景的變遷而損害其原有功能與價值，此種現象值得政府嚴肅看待，並及早提出因應之道。

Since the Agricultural Development Act Amendment has been enacted in 2000, the era of free trade in farmland began and brought great impact on rural landscape. The purpose of this study is to explore rural landscape change in Yilan County and to predict future distribution of farmhouses. Three typical countryside villages Weiqian, Dazhou, and Dayi, of Sanxing Township were selected as the experimental sites. The Markov model was applied to predict future landscape changes, and the Logistic regression model with six factors was used to analyze the spatial correlation between rural farmhouses and factors. Results obtained from these two models were then integrated to simulate the spatial distribution of rural farmhouses. We found that: (1) Changing speed of farmland converting into farmhouses was faster during 2000-2009 than 1987-2000. It indicates agricultural policy influenced landscape change. (2) Farmhouses may convert into farmland and become main land type in 2054 if they continuously develop in the rate of 2000-2009. (3) Simulation maps of 2018, 2036 and 2054 showed that the spatial distribution of farmhouses has the tendency of expansion possibly due to the rapid disappearance of farmlands. Apart from the agricultural production, the maintaining of farmland ecosystem, the supply of open space, and the value of beautiful scenery might be lost with landscape changing. Findings should facilitate decision making for the regional development.