An optimal three-dimensional visualization system was developed for edging and trimming of rough lumber in central Appalachia. ActiveX Data Objects were implemented via MS Visual C++/Open Graphics Library to manipulate board data at the backend supported by a relational data model with four data entity types: Board, Shape, Defect, and Defect Type. Exhaustive search procedures and a dynamic programming algorithm were used to achieve the optimal edging and trimming solution. A lumber grading module was also developed to grade hardwood lumber based on National Hardwood Lumber Association grading rules. The system was validated through comparing the total lumber values generated by the system with those generated by six local sawmills. A total of 360 boards were measured for board dimensions, defects, shape, wane, and the results of edging and trimming. Results indicated that the lumber value and surface measure gained in these six sawmills could be increased, on average, by 21.37 and 6.1 percent, respectively, using the optimal edging and trimming system. The optimal edging and trimming system not only can be used as a training tool but also can be installed on a field PC to aid the edging and trimming process.