Background Sensorimotor impairments of the upper limb (UL) are common after stroke, but there is a lack of evidence‐based interventions to improve functioning of UL. Objective To evaluate (1) the efficacy of sensory relearning and task‐specific training compared to task‐specific training only, and (2) the feasibility of the training in chronic stroke. Design A pilot randomized controlled trial. Setting University hospital outpatient clinic. Participants Twenty‐seven participants (median age; 62 years, 20 men) were randomized to an intervention group (IG; n = 15) or to a control group (CG; n = 12). Intervention Both groups received training twice weekly in 2.5‐hour sessions for 5 weeks. The training in the IG consisted of sensory relearning, task‐specific training, and home training. The training in the CG consisted of task‐specific training. Main Outcome Measures Primary outcome was sensory function (touch thresholds, touch discrimination, light touch, and proprioception). Secondary outcomes were dexterity, ability to use the hand in daily activities, and perceived participation. A blinded assessor conducted the assessments at baseline (T1), post intervention (T2), and at 3‐month follow‐up (T3). Nonparametric analyses and effect‐size calculations (r) were performed. Feasibility was evaluated by a questionnaire. Results After the training, only touch thresholds improved significantly from T1 to T2 (p = .007, r = 0.61) in the IG compared to the CG. Within the IG, significant improvements were found from T1 to T2 regarding use of the hand in daily activities (p = .001, r = 0.96) and movement quality (p = .004, r = 0.85) and from T1 to T3 regarding satisfaction with performance in meaningful activities (p = .004, r = 0.94). The CG significantly improved the performance of using the hand in meaningful activities from T1 to T2 (p = .017, r = 0.86). The training was well tolerated by the participants and performed without any adverse events. Conclusions Combined sensory relearning and task‐specific training may be a promising and feasible intervention to improve UL sensorimotor function after stroke.