The allocation of adjustability and use of discrete sizes are two methods of accounting for the variability in the population of prospective users of a product. This paper addresses the discrete sizing problem, adapting recent adjustability research. This is done in the context of the optimization of tool handle sizes, and a case study of designing multiple sizes of a nonuniform cylindrical tool handle is presented. Specifying multiple sizes of a product introduces additional considerations and complexities compared with specifying adjustability. It is demonstrated that altering performance specifications has a significant impact on the outcome. For example, maximizing average performance yields both the best performance for the greatest number of users and the highest level of cumulative performance for the population. In contrast, optimizing minimum performance ensures that all users experience a baseline of comfort or safety. Measuring the accommodation of a population across multiple sizes is shown to be a two-step process consisting of an assessment of performance for an individual followed by a comparison of that performance with an established standard across the population (i.e., in the current work, a term called grip quality, Q, is assessed for each user in a virtual population, and then all users with Q ≥ 0.95 are considered to be accommodated). Assessing accommodation in this manner facilitates the decision-making process of the designer as they are able to weigh the tradeoff between more sizes and increased cost by finding a point of diminishing marginal performance. The results of the work include specific handle design recommendations for the population considered and a methodology applicable to the design of other products.