The ability of the Cognitive Status Examination (CSE), a brief neuropsychological screening instrument, to detect cognitive impairment in substance abusing patients was evaluated in 200 consecutive admissions to a drug/alcohol rehabilitation unit. While a significant proportion (25%) of patients screened were classified as impaired using the cutoff derived in the original study, agreement with an independent neurological impairment classification was much lower than that reported in the earlier work. Possible reasons for this discrepancy are discussed, including the relatively subtle nature of cognitive impairment in substance abusing patients. Implementation of regular screening did seem to heighten staff sensitivity to cognitive impairment in this sample. Alternate methods of determining the usefulness of cognitive screening tools (e.g., relationship to treatment relevant skills) are also discussed.