A goal of rehabilitation for children with hemiparesis is improved arm use during daily life. However, paediatric outcome measures do not objectively measure arm use in natural environments. Accelerometers, which quantitatively measure movement, may provide some missing information. This thesis examines the criterion, convergent and discriminative validity of accelerometry as a measure of paediatric arm movement. The results from a sample of 26 typically-developing (TD) participants and 13 participants with hemiplegic cerebral palsy (CP) indicate arm-accelerometry has: i) criterion validity, based on its sensitivity, specificity (Youden’s Index, J > .63) and agreement (> 84%), with reference to video-based observations; ii) moderate convergent validity, based on the correlation between arm ratios and QUEST scores (rs = 0.662, p = .026); and iii) discriminative validity, based on differences in arm ratios (p