This paper investigates the relationship between R&D capabilities (learning, R&D and external networking), technology commercialization (manufacturing and marketing), and innovation performance (product competitiveness) among SMEs in IT-related businesses. The study focuses on 254 Korean IT SMEs that were either recipients of government R&D grants or their indirect beneficiaries during the two-year period between 2005 and 2007. The major findings of this study are as follows: First, unlike what has been suggested by previous studies, R&D intensity was not the only factor influencing the innovation performance of firms; learning and external networking also had a significant influence on innovation. The research implication of this finding is that the measurement of firms' performance should not be solely based on the intensity of R&D expenditures, but a broader set of factors including learning and external networking capabilities. Second, the technology commercialization capabilities of firms played the role of a mediator in the relationship between R&D and innovation performance. Within the innovation cycle of input (R&D capabilities), process (technology commercialization capabilities) and output (innovation performance), we found that R&D seldom influenced performance in a direct fashion, but its influence was most often mediated by technology commercialization capabilities. The practical implication of this finding for companies is that in order to improve performance, they must avoid narrowly focusing on R&D, but must invest also in capabilities to commercialize technologies resulting from R&D. Third, when direct and indirect beneficiaries of public R&D funding are compared together, the explanatory power of the relationship between R&D capabilities, technology commercialization capabilities and innovation performance was stronger among the latter than the former. This result suggests that indirect technology support toward Korean IT SMEs through government-sponsored research institutions is a more effective way of allocating public R&D funds than direct funding in the form of grants to individual companies. In other words, sponsoring R&D projects at research organizations with high-quality manpower and equipment and facilities like government research institutions, which are more likely to result in technologies that are readier for commercialization and have greater value-added, and transferring resulting technologies to small and medium-size ventures is a better strategy for enhancing national technological competitiveness in IT.