The validity of self-reported intoxication at time of spinal cord injury (SCI) was examined for 88 cases at admission to an acute SCI center by comparing self-reports with serum and urine analyses. Serum ethanol greater than 50 mg/dl was the most frequently found substance (observed in 40% of the cases) followed by urine analysis evidence of cocaine (14%), cannabinoids (8%), benzodiazipines (5%), and opiates (4%). Evidence of substances with abuse potential was found in urine for 35% of the sample. While 62% of the sample had either serum ethanol greater than 50 mg/dl or a positive urine analysis, only 42% of the sample reported being under the influence of some substance at the time of SCI. Although the relationship between these two measures was statistically significant, self-report and toxicology analyses were discordant in 34% of the cases. These results suggest that routine drug testing at admission to an SCI center will produce both false-negative and false-positive results if substance presence alone is interpreted as evidence of intoxication.