For a long time, Substance Use Disorders (SUDs) were not considered a component in the etiology of dementia. The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders introduced substance-induced neurocognitive disorders, incorporating this notion to clinical practice. However, detection and monitoring of neurodegenerative processes in SUD patients remain a major clinical challenge, especially when early diagnosis is required. In the present study, we aimed to investigate new potential biomarkers of neurodegeneration that could predict cognitive impairment in SUD patients: the circulating concentrations of Neurofilament Light chain protein (NfL) and Brain-Derived Neurotrophic Factor (BDNF). Sixty SUD patients were compared with twenty-seven dementia patients and forty healthy controls. SUD patients were recruited and assessed using the Psychiatric Research Interview for Substance and Mental (PRISM) and a battery of neuropsychological tests, including the Montreal Cognitive Assessment test for evaluation of cognitive impairment. When compared to healthy control subjects, SUD patients showed increases in plasma NfL concentrations and NfL/BDNF ratio, as well as reduced plasma BDNF levels. These changes were remarkable in SUD patients with moderate–severe cognitive impairment, being comparable to those observed in dementia patients. NfL concentrations correlated with executive function and memory cognition in SUD patients. The parameters “age”, “NfL/BDNF ratio”, “first time alcohol use”, “age of onset of alcohol use disorder”, and “length of alcohol use disorder diagnosis” were able to stratify our SUD sample into patients with cognitive impairment from those without cognitive dysfunction with great specificity and sensibility. In conclusion, we propose the combined use of NfL and BDNF (NfL/BDNF ratio) to monitor substance-induced neurocognitive disorder.