Abstract: Sensory impairments in the upper limb are common after stroke. Different sensory modalities can be affected such as sense of touch, pressure, pain, temperature and proprioception. The sensory impairments can negatively affect motor function, dexterity and the ability to perform daily activities. Despite these problems limited attention is paid to sensory impairments in stroke rehabilitation. The overall aim of this thesis was to increase knowledge about the consequences of sensory impairments of the upper limb after stroke, and to evaluate the effects of a novel training approach.The thesis comprises four studies. In study I, 15 participants aged 35-78 years were interviewed individually about their experiences of sensory impairment in the upper limb after stroke. Data were analysed by inductive content analysis. In study II, factors associated with dexterity were evaluated by linear regression models among 75 participants with mild to moderate stroke. Dexterity was the dependent variable and age, gender, affected hand, social situation, vocational situation, grip strength, spasticity, sensory function, and pain were independent variables. In study III, 27 participants were randomized to either sensory relearning in combination with task-specific training (intervention group, n=15) or task-specific training only (control group, n=12). Both groups trained 2.5 hours per day, twice weekly for 5 weeks. Primary outcome was sensory function of the upper limb , and secondary outcomes were dexterity, ability to use the affected hand in daily activities and perceived participation. An independent assessor conducted all the assessments at baseline, post-treatment and at 3 months follow-up. In study IV, the 15 participants who had underwent sensory relearning in combination with task-specific training in study III were interviewed about their experiences and perceived effects.The results showed that an impaired sensory function in the upper limb after stroke had a great impact on personal tasks and on everyday activities and leisure activities. Despite this, specific training to improve sensory function was lacking and had not been a part of the participants´ rehabilitation (study I). Sensory function in terms of discriminative touch was a major contributing factor to dexterity in persons with mild to moderate stroke explaining 46% of the variance in dexterity (study II). After sensory relearning in combination with task-specific training there was a significant between group difference in touch detection (p = 0.007) in favour of the intervention group but not in any other outcomes. A significant improvement (p