Despite the extensive body of knowledge that exists on innovation management and project management, the number of failed projects is still disappointingly high. This is especially true in projects that involve break-through innovations. Recent literature suggests that the application of a contextual approach rather than a universal approach could improve this situation. This observation has prompted a search for factors that affect the performance of projects, as well as the development of several contextual frameworks. This research builds upon these areas and specifically aims investigates the relationships between the context and characteristics of a project, the leadership styles, and the decisions that project leaders make early on in projects. The choice of these fields will be explained individually. It is suffice to say at this point that even though research in each of these topics on its own is rather well-developed, the connections between them have not been fully researched before. As an example: the focus of this thesis is on the early phases of projects - also known as the front end development (FED) phase, because in this phase decisions are not yet finalized. Project costs in this phase are still low and making changes is easier. This thesis uses an exploratory study, due to the unavailability of prior integrated research. In this work, we review the literature and attempt to find the project contexts, project characteristics, and leadership styles that influence a project leaders’ decision making. We also look at the decisions that project leaders actually make during the FED phase. We then seek for plausible relationships among these factors and decisions, and in the process build a framework which includes all these factors and decisions. In our proposed framework, the characteristics and contexts of a project influence the decisions that managers make in the early phases of a project. The chosen leadership styles moderate these relationships. We complement the model by presenting several scenarios that aim to refine and show the plausibility of the model. Based on our study, we propose the following conclusions: first, technological uncertainty has a direct and positive effect on risk management, and yet a direct and negative effect on planning, which means that in the case of high technological uncertainty, risk management should be formalized but planning not. These relationships are positively moderated by the leadership variable of task behavior, so if the project leaders’ style is delegating, this would imply more formality in risk management, and more flexibility on planning. The other major conclusion is that project complexity has a direct and negative effect on both risk management and planning, which means that in cases of highly complex projects, both risk management and planning need to be formalized. These relationships are positively moderated by the variable of task behavior, so if the project leaders’ style is telling this would mean more formality in risk management and planning. The model and conclusions can be seen as forms of working hypothesis. In future work, researchers could focus on testing these hypotheses quantitatively and assess the validity of the model. Eventually, the model can be complemented by adding more contextual variables. The proposed conceptual model connects the fields of innovation management, project management and leadership theory. We hope that the model contributes to these fields and to the understanding of the realities which project leaders face, and that it can act as a stepping stone for further research.