OBJECTIVE: To test a theoretical model examining processes through which a parent-focused educational-behavioral intervention [Creating Opportunities for Parent Empowerment (COPE)] relates to children’s post-hospital adjustment problems. METHODS: Mothers (n = 143) and their 2-7-year-old children, unexpectedly hospitalized in two pediatric intensive care units, were randomized to COPE or control conditions. Maternal measures included parental beliefs, anxiety, negative mood, and child adjustment 3 months post discharge. Observers rated maternal support of their children during hospitalization. RESULTS: Structural equation modeling suggested that the model tested provided a reasonable fit to the data [chi2 (97 df) = 129.43; p = .016; root mean square error of approximation = .048; comparative fit index = .95]. COPE effects on children’s post-hospital externalizing behaviors were indirect, via associations with parental beliefs and maternal negative mood state. Furthermore, COPE participation was associated with more maternal support of their children, which was also associated with less internalizing and externalizing behaviors 3 months post discharge. CONCLUSION: Implementing COPE may help avert future mental health problems in this high risk population. Understanding the processes by which an already empirically validated program relates to child outcomes is likely to aid clinicians and administrators in the widespread uptake of the COPE program.