The paper examines the evidence of R & D spillovers through trade and foreign direct investment at the industrial sectoral level. Specifically, it focuses on the appropriate measure of foreign R&D stocks embodied in imports, and proposes an alternative measure to the traditional measure. With a new measure of foreign technology, this paper then empirically distinguishes between R&D spillovers from import volume and from import variety. Panel estimations with fixed effects use a data set comprised of eight sectors in the manufacturing industry of thirteen OECD countries from 1981 to 1999. This empirical work at disaggregated level has shown the evidence to generally support the role of manufacturing imports in the international diffusion of technology. In particular, the results imply that the impact of import variety proves stronger and more significant in investigating trade-related R&D spillovers. The evidence indicates the variety effect dominates the volume effect, and hence import variety proves more important in explaining trade-related technological diffusion. This paper also investigates R&D spillovers occurring simultaneously through trade and non-trade (i.e., FDI) channels, Empirical results provide no strong evidence on R&D spillovers through FDI, although FDI is often considered as an important channel for technological diffusion. The evidence implies that the effect of trade channel dominates that of the FDI channel and finds that FDI only weakly affects domestic productivity.