Research shows that adult hospital built environments can affect physical and psychological outcomes and healthcare satisfaction, but research on pediatric built environments is sparse. This study investigated the effects of the built environment on pediatric hematology-oncology patients, their parents, and nursing staff at a Southern California children’s hospital serving a predominantly Latino population. Hospital built environment was evaluated subjectively through satisfaction questionnaires and objectively by quantifying environmental features. All subjects completed a questionnaire battery measuring the constructs of present functioning (PF), positive affect (PA), and negative affect (NA). Healthcare satisfaction (HS) was assessed for parents. Co-worker satisfaction (CS) and multidimensional fatigue (F) were assessed for staff. Patient medical data were obtained from charts. Structural equation modeling (SEM) was used to test the hypothesis that environmental satisfaction (ES) would mediate the relationship between objective environment (OE) and PF, PA, and NA (with HS, and CS and F included as outcomes in the parent and staff models, respectively), such that the presence of more environmental features thought to be beneficial would predict greater ES and better outcomes. For less-acculturated Latinos, we hypothesized that ES and outcome measures would be more strongly associated with “visitor facilitating features” (VFFs) than for more acculturated Latinos. Subjects were 90 hematology-oncology patients, 149 parents, and 113 staff. The mediational hypothesis was not supported for children, but partially supported for parents (i.e., OE → ES → HS), and fully supported for staff (i.e., OE → ES → PF, F, NA, PA, CS), with all relationships in the expected directions. Exploratory post-hoc SEMs that maximized statistical power by excluding covariates revealed a significant positive relationship between OE and ES for children, and significant negative relationships between ES and PF, and ES and NA for parents, in predicted directions. Significant relationships opposite than expected emerged between OE and parental PA, and between OE and staff CS. For less acculturated Latino parents, the presence of VFFs was more predictive of greater HS and lower NA, than for more acculturated Latinos. Results of exploratory analyses probing effects of individual environmental features are discussed.