IntroductionEstablishing differential diagnosis between different inflammatory causes of acute transverse myelitis (ATM) can be difficult. The objective of this study was to see which clinical, imaging or laboratory findings best contribute to confirm ATM etiology. MethodsWe reviewed clinical history, MRI images, CSF and serum laboratory tests in a retrospective study of patients presenting ATM. Univariate and multivariate multinomial logistic regression analysis was performed for each of the items listed above. ResultsOne hundred and seventy-two patients were analyzed in the study: 68 with multiple sclerosis (MS), 67 presenting idiopathic myelitis (IM; 23 of which were recurrent), 21 who developed positive systemic-antibodies associated myelitis (SAb-M) and 16 with neuromyelitis optica spectrum disorders (NMOSD). The following factors were associated with increased risk of developing MS: lower values in the modified Rankin scale at admission; positive oligoclonal bands (OCB); higher spinal cord lesion load; presence of brain demyelinating lesions; and disease recurrence. Longitudinally extended (LE) lesions, brain demyelinating lesions, and recurrences also contributed to final diagnosis of NMOSD. Multivariate multinomial logistic regression analysis showed presence of LE lesions increased risk of NMOSD and recurrence of ATM. Whereas, brain demyelinating lesions, and presence of OCB increased risk of MS. ConclusionsATM etiology may be clarified on the basis of spinal cord and brain MRI findings, together with CSF biochemistry and serum laboratory test results, allowing more timely and exact diagnosis as well as specific therapy for cases of uncertain origin.