Shoulder joint stability mediated by proprioception is often quantified by arm repositioning tests (i.e., static end-position accuracy), overlooking ongoing movement quality. This study assessed movement quality by adopting smoothness-related kinematic descriptors. We compared performance of healthy controls with that of patients in arthroscopic shoulder stabilization and open shoulder stabilization groups. We hypothesized that arm kinematics after arthroscopic intervention would more closely resemble healthy movements compared with patients after open shoulder stabilization surgery. Healthy controls (N = 14) were compared with patients after arthroscopic shoulder stabilization (N = 10) and open shoulder stabilization (N = 12). Right-hand dominant subjects (the affected side in patients) performed 135 unconstrained 3-dimensional pointing movements toward visual targets (seen through pinhole goggles; i.e., no arm vision). Arm kinematic data were recorded and offline analyzed to obtain hand tangential velocity profiles further used to compute the acceleration-to-movement time ratio, peak-to-mean velocity ratio, and number of velocity peaks ("symmetry," "proportion," and "fragmentation" features, respectively). Parametric and nonparametric statistics were used for comparisons (P ≤ .05). Control and arthroscopic shoulder stabilization groups presented similar acceleration-to-movement time ratio and peak-to-mean velocity ratio. Both groups differed from the open shoulder stabilization group (P = .001). Distributions of velocity peaks for control and arthroscopic shoulder stabilization groups were similar, whereas open shoulder stabilization and control subjects differed significantly (P = .028). Movement quality mediated by proprioception in arthroscopic shoulder stabilization patients matches that of healthy controls, whereas performance in open shoulder stabilization patients seems inferior compared with that in healthy controls, as assessed by smoothness-related measures (less symmetrical, more fragmented movements).