Substance exposed newborns (SENs) are at risk of developmental delay(s). Early intervention (EI) access, key to addressing these risks, is inequitable. Objectives were to determine prevalence of EI referral in the Colorado Hospitals Substance Exposed Newborn (CHoSEN) Quality Improvement Collaborative (USA) and evaluate predictors of referral. Maternal-infant dyads with exposure to medications for opioid use disorder (MOUD), illicit/prescription opioids, and/or non-opioid substances were included. Chi-square, Fisher’s exact, and ANOVA tests evaluated differences in maternal/infant characteristics by referral. Multivariable logistic regression models assessed the independent association of characteristics with referral. Among 991 primarily White, non-Hispanic dyads, 407 (41.1%) SENs received EI referral. Infants born to mothers with non-MOUD (AOR 0.22, 95% CI 0.10, 0.45) and polysubstance (AOR 0.60, 95% CI 0.36, 0.98) exposure were less likely to receive referral compared to infants born to mothers with MOUD exposure. Those with shorter length of stay of 1-3 days (AOR 0.03, 95% CI 0.01, 0.07) and 4-6 days (AOR 0.47, 95% CI 0.32, 0.69) were less likely to receive referral compared to infants admitted >6 days. Less than half of identified SENs received EI referral with variation by substance exposure and length of stay. Systems to ensure equitable access to services are crucial.