Innovation drives economic growth and connects with the firms evolutionary cycle that launches new products in a continuous quest for performance improvement. With the advent of open innovation, more firms have begun to complement their knowledge base and to fuel the innovative process with external sources. However, when innovation is developed, part of the new knowledge also becomes public through knowledge spillovers, thus potentially benefiting other firms. The relation between innovation performance and external knowledge sources, such as the firm s collaboration with partners, has been extensively studied in developed markets. However, the same is not observed for knowledge spillovers, especially in the case of countries that are less advanced with respect to innovation. This study analyzed the extent to which external knowledge provided by the collaboration with partners and especially the industry knowledge spillovers affect product innovation performance and firm growth. A conceptual model was proposed to measure the relative contribution of these sources as well as the impact of the firm size on the studied relations. The analysis was conducted on a sample of 913 Colombian manufacturing firms with data extracted from EDIT innovation research for the period 2011 to 2016. The hypotheses were tested using the structural equation modeling technique. To understand the impact of firm size on the model relationships, a multigroup analysis was carried out. The results supported most hypotheses and provided some interesting insights. The positive effects of the firm collaboration on product innovation performance were corroborated, revealing that resource-scarce environments, such as Colombia, encourage collaboration regardless of the size of the firm. As for the positive effects of industry knowledge spillovers on product innovation performance, support was contingent on the firm size. For small and medium enterprises (SMEs), not only were these effects confirmed but also proved to be superior to those of collaboration with partners. This finding makes it evident that knowledge spillovers are the most relevant source in explaining the product innovation performance of Colombian SMEs, even when the firm collaboration is also significant. In the case of the large enterprises (LEs), the direct contribution of the industry knowledge spillovers on product innovation performance was not significant, possibly due to their greater internal abundance of resources. For the effect moderated by the absorptive capacity, the contribution was negative, possibly indicating that the knowledge losses generated by outgoing spillovers are detrimental to the innovation performance of these firms. The study also corroborated that the growth of firms of both sizes is positively influenced by their innovation performance. However, the findings for industry knowledge spillovers suggest that Colombian SMEs may be behaving opportunistically, excessively free-riding on R&D investments of LEs which may reduce the incentives for the latter to keep investing in innovation. It is hoped that these findings will contribute to Colombia s efforts to create better conditions for innovation to thrive.