The concept of portfolio is rarely understood and used in R&D management field. The conventional management technique in selecting R&D projects is scoring each projects and choosing projects based on the scores which are determined during evaluation. As a collection of good stocks is not necessarily a good stock portfolio, a collection of good R&D projects is not always a good R&D project portfolio. In this paper, framework and practical technique for constructing a R&D project portfolio are introduced. This technique can be easily applied in private and plublic R&D institutes.