It has long been stated in the literature the effects of collaboration to innovation, especially regarding research and development (R&D) activities. However, these are dynamic empirical fields. Therefore theoretical approaches face constant challenges to understand and explain the new evidences. Due to the limited scope and scale of organizations to search and identify partners with complementary knowledge and resources, and to select those with potential to effectively cooperate for R&D, there is an increasing emergence of agents who provide these services in the market. Called intermediaries or brokers, they influence the interaction among organizations with the common goal of innovation. Still, the literature has reported that the intermediary may play an important set of functions for R&D projects not limited to the search of partners. This thesis is therefore mainly concerned with the influence of innovation intermediaries in the context of collaborative R&D projects, suggesting a conceptual framework on the role of intermediaries. The framework emphasises that R&D collaboration goes beyond dyadic relationships usually highlighted in the literature. The roles of intermediaries provide an important additional dimension in collaborative R&D projects. The empirical part of the thesis explored three case studies: Force for Elastomers, from the Federal University of Rio Grande do Sul in Brazil; the Orange Service Call and Reward project undertaken by the National Endowment for Science, Technology and the Arts (NESTA) for Orange; and the StarStream project from the University of Southampton, both in the United Kingdom. The results confirmed the influence of innovation intermediaries in most of the critical elements of a R&D project. The study advances the understanding of the influence of intermediaries for the beginning of a new project between partners. The analyses also show that intermediaries influence especially through the search for possible partners and the management of the relationship. However, the activities of research and knowledge production as well as activities of development and prototyping were not directly influenced by intermediaries. The stage after R&D, when the partners had reached positive results from activities, received a major influence from intermediaries who helped the firms to protect the inventions and to value and commercialize the new technology. Research outcomes still reveal that there is still a lack of measurements about intermediaries’ effectiveness and therefore firms involved in partnerships cannot fully evaluate their role.