The purpose of this paper is to investigate productivity growth according to three types of innovation: product innovation, product improvement, and process innovation, using firm-level data from the Korea Innovation Survey 2002 about Korean manufacturing companies. This study is conducted in two steps. The first step measures firm-level Malmquist productivity index by data envelopment analysis (DEA). The second step estimates econometric regression models by weighted least square utilizing the productivity index as a dependent variable and each of the innovation types and other characteristics of firms as independent variables. The empirical results partly support the idea that the types of innovation matter in productivity growth. Specifically, process innovation may result in higher productivity performance than product innovation in the short run. This result stems from the difference in efficiency growth when productivity growth is decomposed into two components: efficiency growth and technical growth. That is, product innovation by definition involves product development and radical innovation and so, it can deteriorate efficiency growth relative to other types of innovation due to the process of product development and the adjustments that are needed to new innovations whereas process innovation is implemented to reduce defects, lead time, costs and other factors, and as such is very efficiency orientated. Consequently it helps improve efficiency growth.