In this paper we incorporate R&D capital explicitly into the analysis of total factor productivity growth. The introduction of R&D capital as an input variable in addition to labor and fixed capital tends to reduce our estimate of Malmquist indexes for a sample of 14 OECD countries, on average, by 10 percent. Our findings indicate that it is technological progress rather than efficiency catch-up that is driven by the accumulation of R&D capital. We also find that productivity gains in manufacturing industries depend importantly on R&D spillovers.