INTRODUCTION: Premorbid alcohol and drug use is frequently seen in traumatic brain injury (TBI), yet in-hospital resource utilization in this population has not been thoroughly studied. METHODS: The multicenter TRACK-TBI database was queried for patients with blood alcohol content (BAC, positive ≥0.08), cannabinoid (THC), and hard drug (cocaine, amphetamines, PCP, MDMA) screening at presentation. Patient demographics, presenting GCS, ICU admission, ventilator time, number of head CTs, ICP monitor placement, and length of stay (LOS) were analyzed after stratifying by substance use. RESULTS: Of 2032 TRACK-TBI subjects, 1448 (71.3%) had BAC screening, 891 (43.8%) had THC screening, and 891 (43.8%) had hard drug screening. BAC+ patients (n = 370 [25.6%]) had lower presenting GCS (11.5 ± 4.7 v. 12.7 ± 3.9, p<0.001) as did THC+ patients (n = 136 [15.3%], 10.9 ± 4.7 v. 12.6 ± 4.0, p < 0.001), although hard drug+ patients (n = 82 [9.2%]) did not (p = 0.119). BAC+ and hard drug+ patients had higher positive head CT rates (p < 0.001 and p = 0.034, respectively), although THC+ patients did not (p = 0.248). All substance use groups were more likely to be admitted to ICU but did not have more ICU days on average. Hard drug+ patients were more likely to require ICP monitor placement (27% v. 14%, p = 0.003) despite having similar presenting GCS. BAC+ patients had significantly more CTs performed (p = 0.007). Only hard drug+ patients required more ventilator days (p = 0.04). LOS was significantly longer for THC+ and hard drug+ patients (mean 10.8 v. 8.1 d, p = 0.035 and 11.5 v. 8.2 d, p = 0.017, respectively) but not BAC+ patients. On multivariate analysis, THC+ (aHR = 1.21, 95% CI 1.04-1.40, p = 0.012) and moderate GCS BAC+ patients (aHR = 1.67, 95% CI 1.20-2.31, p = 0.002) had higher number of head CTs performed. CONCLUSIONS: Substance use associated with TBI is correlated with greater ICU admission rates and imaging resource utilization.