This study aims to cluster track and field athletes based on their average seasonal performance. Athletes’ performance measurements are treated as random perturbations of an underlying individual step function with season-specific random intercepts. A hierarchical Dirichlet process is used as a nonparametric prior to induce clustering of the observations across seasons and athletes. By linking clusters across seasons, similarities and differences in performance are identified. Using a real-world longitudinal shot put data set, the method is illustrated.