Recent studies on human upper limb motion highlighted the benefit of dimensionality reduction techniques to extrapolate informative joint patterns. These techniques can simplify the description of upper limb kinematics in physiological conditions, serving as a baseline for the objective assessment of movement alterations, or to be implemented in a robotic joint. However, the successful description of kinematic data requires a proper alignment of the acquisitions to correctly estimate kinematic patterns and their motion variability. Here, we propose a structured methodology to process and analyze upper limb kinematic data, considering time warping and task segmentation to register task execution on a common normalized completion time axis. Functional principal component analysis (fPCA) was used to extract patterns of motion of the wrist joint from the data collected by healthy participants performing activities of daily living. Our results suggest that wrist trajectories can be described as a linear combination of few functional principal components (fPCs). In fact, three fPCs explained more than 85% of the variance of any task. Wrist trajectories in the reaching phase of movement were highly correlated among participants and significantly more than trajectories in the manipulation phase ( [Formula: see text]). These findings may be useful in simplifying the control and design of robotic wrists, and could aid the development of therapies for the early detection of pathological conditions.