Purpose: The upper limb items of the Motor Assessment Scale (MAS) have been shown to be a sensitive, valid and reliable measure of upper limb function for adults following stroke, however the validity and reliability of summing these items into an independent subscale has not yet been evaluated. The stability, internal consistency and construct validity of the upper limb MAS subscale (UL-MAS) was assessed in this study. Method: Twenty-seven inpatients following stroke (mean age = 67 years, range = 40-80) were sampled from an acute, inpatient rehabilitation setting. Patients were evaluated with 'Upper Arm Function', 'Hand Movements', and 'Advanced Hand Activities' items of the MAS by masked physiotherapists who had received standardized training in administration of the MAS. Results: All items were explained by one factor on confirmatory factor analysis and correlated significantly with one another and with the composite (summed total) score. Internal consistency analysis produced a Cronbach's alpha of 0.83 which did not benefit from removal of any items. Conclusions: The acceptable internal consistency score obtained verifies the validity and reliability of using the UL-MAS as an independent scale. This study has also verified the construct validity of the UL-MAS subscale and provides a valuable extension of previous work, which together demonstrates the value of the UL-MAS as a responsive, valid and reliable measure of upper limb function in adults following stroke. The UL-MAS produced a single, composite score that could be interpreted as a total score for upper limb function in this population.