Site quality is one of the main information needed in forest stand management. Site quality classes need to be evaluated every certain period because the quality of forest stands may change as a result of management applied. This study describes the use of very high-resolution imagery derived from unmanned aerial vehicle (UAV) for estimating the site quality of teak (Tectona grandis). The UAV imagery used was taken from 400 m above datum (the average land surface elevation) with ground spatial resolution of 15 cm. Site quality estimation models was built using discriminant analysis. The study found that the best accuracy from discriminant function using multiple variables canopy density (C) and average of crown diameter (D_c ̅) is 60.9%.