BACKGROUND: Somatosensory discrimination training (SDT) consists of a peripheral stimulation of somatosensory receptors followed by the request to discriminate among different stimuli features. Such approach has been adopted to promote motor learning after stroke, but results are still controversial. This systematic review aimed at investigating SDT effects on motor and functional recovery in patients with stroke. METHODS: A systematic literature search was performed using PubMed, EMBASE, CINAHL, and Cochrane Library from inception until January 2025. Randomized controlled trials investigating the effects of SDT delivered alone and/or in combination with other rehabilitative approaches in patients with stroke were included. Two reviewers performed data extraction and assessed methodological quality using the revised Cochrane risk of bias tool. Pooled or un-pooled effects were computed as mean difference or standardized mean difference with 95% confidence interval, while evidence certainty was rated using the GRADE approach. RESULTS: Nine studies (264 participants) were included. Low evidence supported SDT plus task-oriented rehabilitation to improve upper limb function (MD 0.2 points, CI CONCLUSIONS: The addition of SDT to task-oriented rehabilitation led to better motor and functional improvements than task-oriented rehabilitation alone in patients with stroke. However, evidence certainty was very low to low and further high-quality studies are required.