OBJECTIVE The aim of this study was to determine the inter-rater reliability of the modified Medical Research Council (MRC) scale for grading motor function in patients with chronic incomplete spinal cord injury (SCI). METHODS Two neurosurgical residents and 2 faculty members performed motor examinations in 6 chronic incomplete SCI patients for a total of 156 muscle groups. Examinations were performed using the modified MRC grading scale during routine clinic visits for each patient. Informed consent was obtained prior to enrollment. Patients with American Spinal Injury Association (ASIA) Impairment Scale grade A (ASIA A) injuries were excluded. Inter-rater reliability coefficients were calculated using Kendall's coefficient of concordance (W) and intraclass correlation coefficients (ICCs). RESULTS Sixty-four percent of the tested variables demonstrated extremely strong (W 0.71-0.9) or strong (0.51-0.7) inter-rater reliability using Kendall's coefficient of concordance and an ICC corresponding to excellent (ICC > 0.75) or fair to good (ICC 0.4-0.75) inter-rater reliability. An additional 7% showed poor inter-rater reliability (ICC < 0.4). The remaining variables tested did not reach statistical significance. CONCLUSIONS The inter-rater reliability of the modified MRC scale was found to be high in the majority of tested variables, but the results suggest that discrepancy among trained observers does exist. Reliability was greatest in the lower-extremity muscle groups and least in the upper-extremity muscle groups in patients with chronic incomplete SCI.