The main objective of this study was to examine whether employer-sponsored training has significant effects on labor productivity in the auto parts industry in Taiwan. The uniqueness of this study is that it treats the firm, instead of the individual, as the observation unit and utilizes the direct measure-labor productivity for the analysis of training effects. Among the 600 auto parts manufacturing firms, 237 large and medium firms were selected for empirical analysis. A total of 173 firms responded to the survey. Eleven surveys were not usable, yielding an effective return rate of 68.4％. Data were collected through a self-developed survey with the support of the Taiwan Transportation Vehicle Manufacturers` Association. In order to capture different dimensions of training effects on productivity, multiple measures were sued for both the dependent variable (labor productivity) and the study variable (training). Other factors (control variables) that may affect productivity, such as compensation related, and human resources related policies were entered into the model to thest the robustness of the results. Based on this study`s conceptual framework developed from the human capital theory, it is hypothesized that there is a positive relationship between the amount of training and the level of labor productivity. Most findings from this study`s basic models have supported the hypothesis. The results also showed a fairly high level of robustness when the set of control variables was entered into the model.