This research paper will attempt to describe and explain the relationship between corporate R&D investments and the output of this investment i.e. patents. The first section of the thesis describes the empirical relationship between R&D and patents using standard econometric techniques, and then compares these results with results from a nonparametric technique called locally weighted regression that makes only general assumptions regarding the shape of the regression function. Throughout the analysis, careful attention is payed to the specification of lag structure. In terms of in-sample goodnesss-of-fit and out-of-sample forecasting performance, several parametric models perform well and all models are dominated by the nonparametric procedure with suitably chosen smoothing parameters. The second section of the thesis employs standard event study techniques to investigate whether capital markets favourably perceive these innovative activities. Using a wide variety of R&D and new product announcements from the 1980's, it is found that markets do not respond to announcements of R&D investment changes, although they do reward announcements of new products, which can be viewed as the output of successful R&D