本研究主要探討臺灣堡圖面資料的簡括化程度，以宜蘭廳為研究區，比較水、旱田數化面積與史料面積的差異，分析結果指出，二者相差不大，相關係數高於0.96，線性迴歸解釋力高於0.92，但分組分析發現約5甲以下的土地類別變動較大，因此，若研究議題為小面積的土地類別，宜留意簡括化所造成的影響。雖然簡括化會過濾掉一些資料，但透過分析概知其誤差程度，亦能對相關研究有所幫助。

The main focus of this study is on the area data generalization of Taiwan Bau-Tu. We compare the area discrepancy of the rice paddies and dry farmlands between Taiwan Bau-Tu and historical materials of I-Lan County. The results show that the differences are small, where the correlation coefficient reaches as high as 0.96 and the linear regression's coefficient of variation is over 0.92. If the area is analyzed after divided into groups, the land status of those under 5 Jia (almost equal to hectare) has more variation. For research topics on the land status of smaller areas, it is advised to take notice of the effects of generalization. Although generalization will tend to filter out some information, the discussion on the generalization of Taiwan Bau-Tu can let us understand the amount of errors and can do some contributions to correlative researches.