As companies develop more and more collaborative inter-firm relationships with others on new product development, successful configuration and management of R&D alliance portfolios becomes even more challenging. This paper seeks to understand how different characteristics of a firm’s R&D alliance portfolio influence its ability to acquire external knowledge and how knowledge acquisition in turn influences the firm’s innovation outcomes. The central argument is that a firm’s R&D alliance portfolio strategy entails both knowledge structure benefit and relational benefit. Firms seek to achieve these two different types of benefits by bringing in different desired partners. As a result, a firm’s choice or preferences for different types of R&D alliance partners influence both the scope of external knowledge it is exposed to (knowledge structure benefit) as well as the partner’s willingness to share and the focal firm’s ability to absorb and learn from its partners (relational benefit). While certain strategies may offer firms’ the opportunity to achieve both types of benefits simultaneously, other strategies may force firms to choose one benefit over the other. In order to exploit the value of the acquired knowledge acquired from external partners, firms need to further apply this knowledge to come up with innovative products or services. Literature has further classified new product offerings into radical vs. incremental innovations based on their innovativeness and customer benefit increase. Due to the different characteristics and knowledge requirements for radical vs. incremental innovation, we investigate how different dimensions play differential roles in the development process of these two types of innovation. We empirically tested our hypothesis in the context of pharmaceutical industry with a panel data of 64 firms over 15 years. The statistical results largely support our hypothesis.