We investigated changes in functional arm use after retraining for stroke-related somatosensory loss and identified whether such changes are associated with somatosensory discrimination skills.Data were pooled (N = 80) from two randomized controlled trials of somatosensory retraining. We used the Motor Activity Log to measure perceived amount of arm use in daily activities and the Action Research Arm Test to measure performance capacity. Somatosensory discrimination skills were measured using standardized modality-specific measures.Participants' arm use improved after somatosensory retraining (z = -6.80, p < .01). Change in arm use was weakly associated with somatosensation (tactile, β = 0.31, p < .01; proprioception, β = -0.17, p > .05; object recognition, β = 0.13, p < .05).Change in daily arm use was related to a small amount of variance in somatosensory outcomes. Stroke survivors' functional arm use can increase after somatosensory retraining, with change varying among survivors.