Background. Stroke is a leading contributor to upper limb disability. Current assessments of motor ability are at best subjective estimates. Rehabilitation could be more effective when informed by an individual's specific capacities rather than diagnostic observation. This study characterizes sub-movements of interest during functional reach-to-target tasks with endpoint analysis.\newline Methods. Using Fugl-Meyer (FM) scores and reach capacities, stroke survivors were placed in either mild or severe impairment groups. Individual completed sets of reach-to-target tasks over two visits. The averaged reach curve for each subject was analyzed for angular orientation and velocity profile. The Procrustes method of statistical shape analysis was modified to identify reaching sub-movements that were incongruous to control sub-movements.\newline Findings. Reach-to-target curves were investigated for 15 individuals with mild impairment (FM \(52 \pm 9\)) and 14 individuals with severe impairment (FM \(16 \pm 7\)). The endpoint kinematic analysis found that individuals with severe impairment demonstrate lower peak velocities (centimeters per second) [\(28.5 \pm 17.3 , 80.3 \pm 29.5\)] which occurred sooner in the movement (milliseconds) [\(36.81 \pm 27.0, 59.6 \pm 14.8\)]. The mild group demonstrated variability in orientation in the medial-lateral space, and movement initiation that was congruous to healthy movement. The severe impairment group demonstrated variability in orientation in every axis and showed incongruous initiation and completion behaviors, with some comparable movement in between. \newline Interpretation. Endpoint analysis is a valuable option for clinicians to inform physiotherapy. In addition to kinematic metrics such as orientation and velocity, sub-movements that are comparable to healthy movement can be identified in both mild and severe impairment.