Background: The lack of consensus on the best methodology for identifying cases of non-traumatic spinal cord dysfunction (NTSCD) in administrative health data limits the ability to determine the burden of disease and provide evidence-informed services. Objective: The purpose of this study is to develop an algorithm for identifying cases of NTSCD with Canadian health administrative databases using a case-based approach. Method: Data were provided by the Canadian Institute for Health Information that included all acute care hospital and day surgery (Discharge Abstract Database), ambulatory (National Ambulatory Care Reporting System), and inpatient rehabilitation records (National Rehabilitation Reporting System) of patients with neurological impairment (paraplegia, tetraplegia, and cauda equina syndrome) between April 1, 2004 and March 31, 2011. The approach to identify cases of NTSCD involved using a combination of diagnostic codes for neurological impairment and NTSCD etiology. Results: Of the initial cohort of 23,703 patients with neurological impairment, we classified 6,362 as the "most likely NTSCD" group (had a most responsible diagnosis or pre-existing diagnosis of NTSCD and diagnosis of neurological impairment); 2,777 as "probable NTSCD" defined as having a secondary diagnosis of NTSCD, and 11,179 as "possible NTSCD" who had no NTSCD etiology diagnoses but neurological impairment codes. Conclusion: The proposed algorithm identifies an inpatient NTSCD cohort that is limited to patients with significant paralysis. This feasibility study is the first in a series of 3 that has the potential to inform future research initiatives to accurately determine the incidence and prevalence of NTSCD.