This paper investigates how internal and external factors affect the choice between alliances and joint ventures (A&JVs) and mergers and acquisitions (M&As) for the external sourcing of research and development (R&D) activities, and whether or not such a choice is really contingent, that is, is it the best choice in terms of its impact on firms' innovative performance under those circumstances? We build a set of hypotheses based on both the transaction-cost theory and the resource-based view, and test them through a secondary data source analysis. We found that companies adopt either R&D M&A or A&JV depending on internal (e.g. resources and capabilities, innovation experience) and external (e.g. degree of industry specialisation) factors. Surprisingly, this contingent choice turns to be effective on innovative performance only for the internal factors, rather than the external. This paper contributes to inter-firm relationships literature by presenting the real advantages of using integrated and contingency theoretical models to understand contingent decisions.