The search for clinically relevant molecular markers in classical Hodgkin lymphoma (cHL) is hampered by the histopathological complexity of the disease, resulting from the admixture of a small number of neoplastic Hodgkin and Reed-Sternberg (H-RS) cells with an abundant and heterogeneous microenvironment. In this study, we evaluated gene expression profiles of 11 selected genes previously proposed as a molecular score for adult cHL, aiming to validate its application in the pediatric setting. Assays were performed by RT-qPCR from formalin-fixed paraffin-embedded (FFPE) lymph nodes in 80 patients with cHL. Selected genes were associated with cell cycle (CENPF, CDK1, CCNA2, CCNE2, and HMMR), apoptosis (BCL2, BCL2L1, and CASP3), and monocytes/macrophages (LYZ and STAT1). Despite using controlled preanalytical and analytical strategies, we were not able to validate the 11-gene score to be applied in pediatric cHL. Principal component analysis (PCA) disclosed 3 components that accounted for 65.7% of the total variability. The second PC included microenvironment and apoptosis genes, from which CASP3 expression was associated with a short time of progression-free survival, which impact was maintained in the unfavorable risk group, Epstein-Barr virus-negative cases, and multivariate analysis (P < .05). Because this is a counterintuitive association, CASP3 active expression was assessed at the protein level in H-RS cells by double immunohistochemistry. In contrast to the association of mRNA levels with a poor therapeutic response, a high number of cleaved CASP3+ cells were associated with longer progression-free survival (P = .03) and overall survival (P = .002). Our results demonstrate the feasibility of using FFPE samples as RNA source for molecular prognostication, but argue against the concept of direct and wide applicability of molecular scores in cHL. We reinforce the potential of CASP3 as an interesting target to be explored in adult and pediatric cHL, and alert for its dual biological role in H-RS cells and tumor microenvironment.