CONTEXT/OBJECTIVE To verify the hypothesis that motor levels (ML) inferred from sensory levels in the upper cervical segments C2-C4 according to the current version of the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) are counterintuitive in cases where the most rostral myotomes C5 and C6 are graded as intact. DESIGN Prospective cohort study of ISNCSCI instructional course participants completing a post-test after the workshop to determine the MLs in two variants of a complete, high cervical spinal cord injury (SCI) case scenario. Both variants were based on the same ISNCSCI sensory and MLs of C2. In the first variant myotomes C5 and C6 were bilaterally graded as intact, while in variant 2 only active movements against gravity were possible (grade 3). SETTING Eight ISNCSCI instructional courses conducted during the study period from November 2012 until March 2015 in the framework of the European Multicenter Study on Human Spinal Cord Injury (EMSCI- http//emsci.org ). PARTICIPANTS Ninety-two clinicians from twenty-two SCI centers. Most of the attendees were physicians (58.7%) or physical therapists (33.7%) and had less than one year (44.6%) experience in SCI medicine. INTERVENTIONS Not applicable. OUTCOME MEASURE The classification performance described as percentage of correctly determined MLs by the clinicians. RESULTS Variant 2 (89.13%) was significantly (P < 0.0001) better classified than variant 1 (65.76%). In variant 1 with intact myotomes at C5 and C6, C6 was incorrectly classified as the ML by the clinicians in 33.15% of all cases, whereas in variant 2 with non-intact C5 / C6 myotomes, C6 was rarely chosen (2.17%). CONCLUSIONS Sensory level deferred MLs in the high cervical region of C2-C4 are counterintuitive whenever the most rostral cervical myotomes are intact. An adjustment of the ML definition in ISNCSCI may be needed.