While Nintendo develops many video games internally, Apple is only marginally involved in the development of iPhone applications. This paper addresses the question: to what extent and how should system firms be involved in the development of complementary products for their "core products"? These core products may include video consoles, electronic book readers, etc. This is a highly relevant question, because the success of core products depends strongly on their complementary products, e.g., video games and electronic books. This study proposes a dynamic model for the degree that system firms should be involved in complementary product development by considering, as contingency factors, the novelty of the core, and the novelty of the complementary products. Both novelties influence the degree to which system firms should be involved in the development of complementary products. In terms of a system firm's involvement, this study makes a distinction between the degree of integration and the degree of ownership. The degree of integration reflects the extent to which the system firm is actively involved in the coordination of the complementary product, e.g., to ensure that the complementary product is optimally aligned with its core product and that the full potential of the core product is achieved. The degree of ownership reflects the extent to which the system firm finances the development of the complementary product and therefore, the degree of formal control authority over the complementary product. This model was tested using data from a survey of 99 development projects for mobile telecommunications applications. The results reveal that integration by the system firm contributes to the performance of complementary products for new core products but has a negative effect if a new complement is developed for a mature core product. In addition, ownership contributes to performance if both the core and complementary products are new. In other circumstances, a clear effect of ownership by the system firm on performance is not found. The implications of our findings are that system firms should be strongly involved in complementary product development when they introduce a new core product, and even more so if the complementary products are new. However, they should decrease their degree of involvement over time as the core product matures and, again, even more so if the complementary product is new. The paper concludes by providing practical implications for system firms in the mobile telecommunications industry and beyond.