Today, sustainability emphasis is imperative for modern design and manufacturing enterprises; therefore, sustainability assessment has attracted increasing attention. In this study, we use cost, environmental impact, and labor time as indicators to measure a comprehensive set of sustainability dimensions: economic, environmental, and social sustainability. Similarly, modular product design (MPD), due to its benefits to design and manufacturing, has been widely accepted as a useful strategy. The need for customizing MPD for various sustainability purposes across a product's life cycle motivates this study. Our research goal is to develop a MPD approach to improve the product life cycle performance for dimensions of sustainability. We also concentrate on two current product design focused research gaps: (1) how to best handle key components and (2) taking into account life cycle uncertainty at the component or product end-of-life (EOL) stage. Key components represent core techniques and can have the highest sustainability impact. We specify key components and then develop three sustainable clustering algorithms to generate component modules based on the identified key components. For EOL uncertainty, we adopt fuzzy logic to assess 14 EOL characteristics and use modified transfer equations to convert fuzzy evaluations to designer's perception toward sustainability. Left-right defuzzification method is employed to estimate the probability of each EOL strategy and calculate the expected EOL sustainability values for each component. The Module Structure Sustainability Index (MSSI) advances toward optimizing sustainability in order to determine the best component modules. A coffee maker case study is used to illustrate the proposed methodology.