In patients suffering isolated head trauma, we have previously shown that levels of circulating catecholamines obtained within 48 hours of trauma correlate with the severity of brain injury and predict outcome and that intoxication blunts this response. The effects of alcohol on the increase in catecholamines in systematically injured patients, however, have not been well defined. From 1983 to 1990, 78 patients (74% male; median age 30 years) with blunt head and multisystem injury, who also had alcohol levels measured within 5 hours of injury, were studied. Norepinephrine and epinephrine levels were assayed by a radioenzymatic technique. Injury severity was assessed by the admission Glasgow Coma Scale (GCS) score (4-15; median, 12), the Injury Severity Score (ISS) (13-50; median, 25) and the volume of blood products administered within the first 24 hours (0-14.4 L; median, 0.5 L). The impact of alcohol on the norepinephrine response to injury was analyzed using multiple linear regression models, including polynomial interaction terms. Norepinephrine levels significantly (p less than 0.0001) correlated with the GCS score and ISS. However, alcohol significantly lowered the norepinephrine response to decreasing GCS score (R = 0.49, p less than 0.002) and to increasing ISS (R = 0.51, p less than 0.0006). The blunting of the catecholamine response was most marked in those severely injured. The rise in norepinephrine concentrations seen with increasing volume of blood replacement was not affected by intoxication. An association between injury severity and epinephrine levels was also present, but not as consistently. Epinephrine concentrations rose with falling GCS score and with increasing ISS values, but unlike norepinephrine, there were no apparent effects of alcohol on changes in epinephrine levels. Thus, in patients suffering head and multisystem injury, catecholamine changes reflect the severity of injury using three different scalers. Furthermore, intoxication blunts only the norepinephrine component of this important biologic response.