As pressure grows on developed nations to move away from fossil fuel-based energy sources, so does the potential for nuclear energy to make its resurgence. However, the complex nature of the design process in nuclear engineering and a regulatory culture of ever-increasing safety standards create unique challenges to the nuclear industry. As in many engineering disciplines, the question is one of trade-offs between safety, performance, cost, and time required to develop the design from paper to real life operation. The possibilities facing a designer are virtually unlimited, with fuel choice, layout and operating conditions just three of the many categories which interact with one another in a highly non-linear manner, making it difficult to quantitatively define these trade-offs. Deciding upon an ‘optimal’ design is therefore traditionally done through expert judgement and an iterative design process. Mathematical optimization methods offer a more formal way to optimize designs by employing algorithms to explore the myriad of possibilities in a methodical manner which can yield increased performance over expert designs. In this thesis, an extensive review of the literature revealed gaps which present opportunities for novel research. Two new algorithms are created with the ability to solve optimization problems with multiple objectives simultaneously without requiring weighting or bias from the designer. They are then applied to a series of problems drawn from both the literature and real world designs. The results demonstrate the algorithms’ effectiveness and robustness as well as their ability to handle complex multi-physics problems with reasonably low computational requirements. This research offers an original and effective tool for performing optimization on nuclear fuel assembly design problems and has advanced the state of the art in both multi-objective optimization and its application to the nuclear engineering industry.