Recently, satellite image data has been widely used in various fields. Extraction of forest information is the field which has great application possibility. Forest type information can be used in forecasting and simulating soil erosion, landslides and debris flow as an important influence factor. Application of satellite imagery has advantages to obtain latest forest type information. NDVI calculated from near infra red band of satellite imagery represents vegetation activity, with seasonal difference of it forest type can be classified. In this study, we analyzed MODIS imagery from 2011 years to 2013 years in Gangwon-do where forest area is large and classified forest type with coniferous and deciduous forests. As a result, there was high correlation between forest type of digital map and it classified from NDVI. The possibility for extraction of forest type information from satellite imagery was confirmed.