Current U.S. law mandates notification of substance-affected infants to state child welfare systems (CWS) for the purpose of assessment and service connections. Yet, CWS appears to identify a small share of substance-affected infants at birth, potentially reflecting inconsistent diagnosis of prenatal substance exposure (PSE) and emerging efforts to limit toxicology testing. This study uses linked maternal and child administrative records (2015 to 2021) for Pennsylvania Medicaid-covered births with CWS cases by age 3 (N = 31,913) to investigate PSE prevalence. Using indicators from CWS records and both child and maternal Medicaid claims, we estimate that 45% of children involved with CWS by age 3 experienced PSE, as well as 59% of children entering foster care. Nearly half of these children were not involved with CWS at birth and a majority did not receive a medical diagnosis of PSE. Yet, extrapolated estimates suggest that 70% of Medicaid-covered children with a PSE indicator will experience a CWS case by age 3. Absent high uptake of robust supportive interventions from other systems or agencies, avoiding notification of PSE at birth may not reduce long-term involvement with CWS among children with PSE.