Introduction We aimed to validate previously derived clinical criteria to predict successful prehospital response to naloxone in patients with altered mental status treated by EMS. We hypothesized that prehospital naloxone criteria would have high sensitivity for effective antidote response, but would be underutilized, in patients with drug-related altered mental status (DRAMS). Methods This study was a secondary data analysis of a prospective cohort of acute DRAMS at an urban ED. Naloxone criteria (respiratory rate (RR) <12, miotic pupils, or drug paraphernalia) and mental status, graded by either AVPU (Alert, Verbal, Painful, Unresponsive) or Glasgow Coma Scales, were abstracted from prehospital care reports. Interventions were compared for effective antidote response (EAR), defined as immediate improvement in RR, AVPU, or GCS. Results EMS transported 249 DRAMS over 17 months (48 % males, mean age 41.5, ALS 33.7 %). Forty-three (17 %) patients met naloxone criteria, of whom 44.2 % received the antidote. Naloxone criteria significantly predicted EAR (OR 7.0, p < 0.05) with 83 % sensitivity (95 % CI, 55-95 %). Miotic pupils (OR 20.0, p < 0.01) outperformed RR (OR 2.3, p = NS) as the best single criterion with 91 % sensitivity (95 % CI, 62-98 %). Conclusions This study validates prehospital criteria to guide naloxone administration. In addition, prehospital naloxone was underutilized for DRAMS. Further studies should address potential barriers to prehospital naloxone administration.