BACKGROUND: Robotic and non-robotic training devices are increasingly being used in the rehabilitation of upper limb function in subjects with neurological disorders. As well as being used for training such devices can also provide ongoing assessments during the training sessions. Therefore, it is mandatory to understand the reliability and validity of such measurements when used in a clinical setting. The aim of this study was to evaluate the reliability of movement measures as assessed in the Armeo Spring system for the eventual application to the rehabilitation of patients suffering from cervical spinal cord injury (SCI). METHODS: Reliability (intra- and inter-rater reliability) of the movement workspace (representing multiple ranges of movement) and the influence of varying seating conditions (5 different chair conditions) was assessed in twenty control subjects. In eight patients with cervical SCI the test-retest reliability (tested twice on the same day by the same rater) was assessed as well as a correlation of the movement workspace to retrieve self-care items as scored by the spinal cord independence measure (SCIM 3). RESULTS: Analysis of workspace measures in control subjects revealed intra-class correlation coefficients (ICC) ranging from 0.747 to 0.837 for the intra-rater reliability and from 0.661 to 0.855 for the inter-rater reliability. Test-retest analysis in SCI patients showed a similar high reliability with ICC = 0.858. Also the reliability of the movement workspace between different seating conditions was good with ICCs ranging from 0.844 to 0.915. The movement workspace correlated significantly with the SCIM3 self-care items (p < 0.05, rho = 0.72). CONCLUSION: The upper limb movement workspace measures assessed in the Armeo Spring device revealed fair to good clinical reliability. These findings suggest that measures retrieved from such a training device can be used to monitor changes in upper limb function over time. The correlation between the workspace measures and SCIM3 self-care items indicates that such measures might also be valuable to document the progress of clinical rehabilitation, however further detailed studies are required.