Many engineering-to-order (ETO) organizations are multi-project capacity-driven production systems in which capacity planning is of major importance in the order acceptance phase. The academic literature, in this area, presents a research-practice gap with a lack of studies on the application of decision support tools to address capacity planning problems in real-world ETO settings. Within this context, the goal of this thesis is to develop a tactical capacity planning solution to support the order acceptance phase of a real-world multi-project organization that produces customised equipments on the basis of ETO policy. This research study lays in the development of mixed integer linear programming models and their practical application to solve production planning problems in the studied organization. As for the theoretical contributions of this thesis, first a deterministic model is presented in which modelling issues that are either not entirely explored in other studies or that have to be adapted to the specificities of the studied setting are taken into account. Moreover, a robust optimization model extends the former model by considering uncertainties of the planning problem. The models were fed with real-world data and solved in order to check whether they actually reflect the planning problem. Furthermore, alternative scenarios were also generated to assist the management board in the order acceptance phase. As for practical implications, for the company s manufacturing planning team, the proposed solution enhanced the decision-making process regarding tactical capacity planning, addressing different shortcomings of the company s current planning method. Empirical results suggest that with a slight increase in cost (0.02 percent) a part component should be processed in-house instead of being outsourced and that with a 0.8 percent increas in cost (which includes hiring 21 percent more personnel) the probability of violating the production plans decreases from 90 percent to 15 percent, representing a much more stable (protected against uncertainty) situation. From an academic perspective, this research adds empirical evidence to enrich the existing literature, as it not only presents a real case application, but also highlights issues that must be considered and managed in a real-world context in order to develop and implement appropriate techniques to cope with the aforementioned planning problem.