Prediction of malignant behavior of pheochromocytoma (PC) or paraganglioma (PG) is of limited value. The Cancer Genome Atlas (TCGA) and the French 'Cortico et Médullosurrénale: les Tumeurs Endocrines' (COMETE) network in Paris (France) facilitate accurate differentiation of malignant PC/PG based on genetic information. Therefore, the objective of this transcriptome analysis is to identify the prognostic genes underlying the differentiation of malignant PC/PG in the TCGA and COMETE databases. TCGA carries data pertaining to multigenomic analysis of 173 PC/PG surgical resection samples while the COMETE cohort contains data involving 188 PC/PG surgical resection samples. Clinical information and mRNA expression datasets were downloaded from TCGA and COMETE databases. Based on eligibility criteria, 58 of 173 PC/PG samples in TCGA and 171 of 188 PC/PG samples collected by the COMETE network were selected. Using Ingenuity Pathway Analysis, the mRNA expression of malignant and benign PC/PG was compared. The 58 samples in TCGA included 11 malignant and 47 benign cases. Among the 171 samples obtained from the COMETE cohort, 19 were malignant and 152 were benign. A comparative analysis of the mRNA expression data of the two databases revealed that 11 up/downregulated pathways involved in malignant PC/PG were related to cancer signaling, metabolic alteration, and prominent mitosis, whereas 6 upregulated genes and 1 downregulated gene were significantly enriched in the functional annotation pathways. The TCGA and COMETE databases showed differences in mRNA expression associated with malignant and benign PC/PG. Improved recognition of prognostic genes facilitates the diagnosis and treatment of PC/PG.