The academic interest in the relationship between Internationalization and Performance (RIP) has grown steadily during the last four decades. Based on different theories and methodological approaches, researchers propose and confirm empirically RIP of various natures. Consequently, they have developed various alternative approaches to model RIP such as negative, positive and linear, positive and negative simultaneously, indeterminate, U standard and U inverted. In this context, the three-stage model of Lu and Beamish (2004) and Contractor, Kundu and Hsu (2003), emerges as an integrator model capable of reconciling other efforts to model RIP. However, this integrator model has, in our point of view, two limitations: (1) the absence of the context of internationalization, in particular, the cultural and institutional environment in the theoretical analysis and development; (2) the inability to show the nature of the impact of dispersion of internationalization on performance. To overcome these limitations, we propose, in this work, to use the dispersion of internationalization as a vector by which the context of internationalization is incorporated into the analysis on the impact of degree of internationalization on performance. This research, based on a sample of 69 large international French enterprises over 7 years, 2001-2007, contributes to RIP research by providing the following new theoretical, empirical and methodological elements: (i) the three-stage model applies not only to the relationships between the breadth of internationalization and performance, or the depth of internationalization and performance, but also to the dispersion of internationalization and performance; (ii) it is not only the degree of internationalization itself that explains the firm’s performance but also the context of internationalization.
