Background: Measuring movement performance in people with neurological damage requires a tool that reflects physiotherapy assessment and clinical reasoning. The Leeds Movement Performance Index (LMPI) was previously developed by a group of neurological physiotherapists to fulfill these requirements. Objective: To assess the reliability of the LMPI for use in neurological physiotherapy practice. Methods: Twelve senior neurological physiotherapists were trained to use the LMPI and then asked to measure the movement performance of five patients whose movement had been previously video-recorded for this purpose. A retest session was completed after two weeks. Data were analysed to establish internal and external reliability. Results: Internal reliability was assessed using Cronbach's alpha coefficient, applied to the entire scale (0.862) and to each item (range 0.795-0.892). External (inter-rater) reliability was assessed by a calculation of the intraclass correlation coefficient for scores awarded by multiple raters (0.959), with individual item reliability ranging from 0.874 to 0.968. External (test-retest) reliability was assessed by calculating the Spearman's rank correlation coefficient between scores obtained on two testing occasions (0.792) with values of individual items ranging from 0.397 to 0.674. A variance components analysis partitioned variance into components arising from between-patient variability (55.2%) between-therapist variability (7.8%) and between-testing variability (2.8%). Conclusions: Results indicate that the LMPI is a reliable measurement tool when used by senior neurological physiotherapists.