Original equipment manufacturers (buyers) are increasingly involving suppliers in new product development as a means to increase efficiency and expand capabilities. To realize such benefits, however, the two firms need to have appropriate communication and goal structures to minimize friction while maximizing design quality. In addition, the effectiveness of the inter-firm interaction process, i.e. their collaboration quality, is also a key success factor. This study draws from Information Process Theory to propose that higher technical and relational uncertainty requires more inter-firm communication. The misalignment between communication intensity and uncertainty reduces both design quality and design efficiency. Goal incongruence, which always lowers project performance, is less harmful for projects with high technical uncertainty due to the potential of the conflict resolving process in improving decision quality and efficiency. Finally I use Hackman's theory of work group effectiveness to propose that collaboration quality fully mediates the effects of communication intensity and goal congruence on project outcomes. The study used an empirical survey of manufacturers as the primary method of data collection. Manufacturers that integrate and assemble complex and discrete products are the target population. Design engineers and project managers from manufacturers were my target respondents. Both SEM and hierarchical regression were used to test the conceptual model. The dissertation made five theoretical contributions. First, I introduced the concept that there is an optimal level of inter-firm communication intensity, exceeding which lowers design efficiency without improving design quality. Second, I theoretically defined and empirically operationalized two types of uncertainty, one on the project level and one on the inter-firm level, which were shown to moderate the effects of inter-firm communication and goal structures on collaboration outcomes. Third, this study examined the conditions when goal congruence is more effective in improving collaboration outcomes. Fourth, this study nominally and operationally defined collaboration quality, a theoretical construct which measure the effectiveness of inter-partner interactions rather than mere existence or amount of certain activities pursued by partners. Finally, I proposed several enhancements to existing construct measures.