Designing sustainable diets within planetary boundaries requires considering not only choosing what foods we eat, but also how foods are planned, purchased, and combined into meals. This thesis examines how optimization-based meal planning can contribute to sustainable diets by reducing food waste and increasing the consumption of locally in-season foods. Key drivers of household food waste are poor meal planning and discrete package sizes. A meal planning model is developed to generate meal plans and shopping lists that account for retail package sizes. Results show that healthy meal plans with negligible food waste are feasible, but that reducing waste by weight alone does not necessarily lower dietary carbon footprints. By incorporating uncertainty in household food needs, the analysis highlights how optimal planning horizons and shopping frequencies vary across household situations. The thesis further examines the implications of shifting towards more local and seasonal diets, given that food miles account for approximately 20% of global food-related greenhouse gas emissions. The results show that meal plans can increase the use of locally in-season foods at lower cost and carbon footprint while maintaining nutritional adequacy throughout the year. Finally, computational challenges arising from mutually exclusive choices in meal planning models are addressed through improved branching strategies. Overall, this research demonstrates that optimization-based meal planning can contribute to the development of more sustainable diets by revealing key trade-offs across nutritional, environmental, and economic dimensions. The findings reinforce existing evidence on the need to shift towards more plant-based diets, while showing how different planning and shopping strategies influence sustainability outcomes.