Modular product design has become a core element aspect of sustainable product design, particularly design for product retirement. This work explores the relationship between product modularity and product retirement costs. Previous statements tying increased modularity to decreased costs, specifically product retirement costs, motivated this work. Modular architecture is traditionally made up of functionally independent clusters of components. Past definitions of modularity have centered on a one-to-one correspondence between form and function. An expanded definition of product modularity has been used in this work, which not only includes function, but also form and life-cycle process (manufacture, assembly, retirement, etc.) relationships. The benefits of modularity with respect to product functionality, product development, production, the supply chain, and other life-cycle elements have been expounded by many works in several fields. Specific advantages associated with modular products include economies of scale, standardisation of assemblies, minimisation of assembly time, and improved serviceability. Beyond these advantages, understanding cost's relationships with modularity is essential, since cost is a major factor in the success of a product. No research has been done to prove the existence of a relationship between modularity and cost. Yet, many projects have been based on this unproven assumption. In this paper, an existing modular design method was focused on product retirement. The method was used to measure and redesign four consumer products and to measure ten more consumer products. Costs associated with retirement including, recycling, reuse, remanufacturing, and disposal were measured for each product and each redesign. Statistical analysis of the data was carried out to look at the relationships between relative modularity, retirement cost, and number of design changes made. As was the hypothesis of this work, as retirement modularity increases, product retirement costs tend to decrease. However, this trend is not statistically significant. Specific products did exhibit significant relationships but the strong modularity – cost tie assumed in the literature is clearly not valid. Our study of this hypothesis was complete but limited in scope. We have begun follow on research that expands this work to additional products and additional life-cycle stages.