Complex regional pain syndrome (CRPS) is a chronic neuropathic pain disorder characterized by both central and peripheral symptoms that can be debilitating in children. CRPS treatment typically consists of intensive physical, occupational, and psychological therapy with evidence supporting the efficacy of this approach. Among these outcomes, some patients report significant improvements in pain while others report no change. Identifying baseline predictors of treatment resistance would refine our treatment approach and provide additional targets for intervention. The current study examined baseline brain structure via cortical thickness and gray matter volume (GMV) in 29 pediatric CRPS patients enrolled in an intensive pain rehabilitation program. All participants underwent MRI using a high-resolution T1-weighted sequence. Patients were categorized as pain treatment “responders” (n=19) or “non-responders” (n=10) based on change in reported pain levels from admission to follow up. Compared to treatment responders, non-responders demonstrated significantly less GMV in the bilateral nucleus accumbens p<0.05 and right: putamen p<0.01, pallidum p<0.05, and amygdala p<0.05. Furthermore, treatment non-responders exhibited significant cortical thickening in the left anterior insular cortex and medial frontal gyrus, and cortical thinning in the bilateral precentral gyrus and superior frontal gyrus; right: middle frontal gyrus, fusiform gyrus, inferior temporal gyrus, middle temporal gyrus, and anterior prefrontal cortex; and left: parahippocampal gyrus. Though we did see significant thinning of the primary motor cortex in treatment non-responders compared to responders, the majority of our findings were localized to regions associated with reward, motivation, learning, and emotion. We, therefore, postulate that treatment non-responders, when compared to responders, likely have an intrinsically reduced reward responsiveness, diminished motivation, and impaired learning, overall contributing to their negative treatment outcomes and chronification of pain. In conclusion, these baseline differences overall suggest these regional morphometric alterations may potentially serve as predictors of treatment response in pediatric CRPS. Furthermore, these areas may also indicate possible targets for future treatment.