Anecdotal evidence suggests that recent rises in foster care caseloads are due to increasing parental substance use, particularly misuse of opioids. This study tests the association between child welfare caseloads and two measures of substance use prevalence, including drug overdose deaths and drug-related hospitalizations, for most counties in the United States over the 2011 to 2016 period. We use several statistical models to account for confounding factors, including models examining cross-county variation as well as within county variation. We find four primary findings. First, both measures positively associate with higher foster care entry rates, after accounting for various demographic, socioeconomic, and other local and state-level factors. Second, we find that the relationship is stronger in metropolitan counties than in non-metropolitan counties. Third, hospitalization rates due to different classes of substances, including opioids, have comparable relationships with foster care entry rates, though alcohol has the strongest relationship. Finally, higher substance use prevalence predicts more complex and severe cases of child maltreatment, with more children ending up in foster care in counties with higher overdose death and drug hospitalization rates.