Ecological Classification System (ECS) is a hierarchical system of ecosystem classification, which stratifies land into ecological units (such as Ecological Land Type, ELT) by using Digital Elevation Models (DEM). However, DEM data with different scales may strongly affect the geometrical precision in the classification. More importantly, no DEM data are readily available for remote areas in China. In this study, we derived one DEM data set from topographic maps with scale of 1 ∶ 10000, and buy another set with scale of 1 ∶ 50000. Through comparing the two DEMs in a typical area of mountainous region in eastern Liaoning Province, the DEM with a scale of 1 ∶ 10000 was considered more suitable to create ELT maps. In general, for the purpose of satisfying requirements of spatial scales and geometrical precision, the most effective method to derive DEM data is to digitize the topographic map. However, development of the technique in deriving DEM data is somewhat still needed on a case-by-case basis. Therefore, in this study, by using the method of experiment comparisons, the key technique of deriving DEM data by digitizing topographic maps was figured out, such as the technique of finding errors and the methods of establishing DEM data. The results indicated that, errors in vector data could be quickly found, and a high precise DEM data was derived by using this newly developed technique.This DEM data can be used to derive Ecological Classification System maps of mountainous region in eastern Liaoning Province. The approach discussed in this paper help the local forestry save 78.44% in the costs. The DEM data created played an important role in accurate decision-making for local forest management activities.