Bladder cancer (BC) is a life-threatening malignancy with high mortality rates. Current prognostic models are insufficient in accurately predicting clinical outcomes, impeding personalized treatment strategies. This study aimed to identify BC subtypes and prognostic gene sets by analyzing changes in immune and hallmark gene sets activity in tumor and adjacent non-tumor tissues to enhance patient outcomes. Utilizing data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO), gene set variation analysis (GSVA) was applied to C7 immune-related and hallmark gene sets from the Molecular Signatures Database (MSigDB). The CancerSubtype R package was utilized for clustering these gene sets into three categories, from which 109 candidate sets were identified using Venn diagrams. A refined subset of seven gene sets was selected through least absolute shrinkage and selection operator (LASSO) regression for the construction of a risk model. Model validity was confirmed with receiver operating characteristic (ROC) and calibration curves, and a nomogram was constructed to integrate risk scores with clinical parameters. Finally, genes from the gene sets of the model were acquired and analyzed for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment and protein-protein interactions (PPI) via plugin Molecular Complex Detection (MCODE) and Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) in Cytoscape in both tumor and non-tumor tissues. Three BC subtypes were characterized by immunologic and hallmark gene sets, with subtype 1 patients showing worse survival. The prognostic model, based on seven gene sets, effectively stratified risk, with high-risk patients having significantly shorter survival. GO, KEGG, and PPI analyses indicated distinct influences of non-tumor and tumor tissues on the prognosis of BC patients. We constructed and validated a novel prognostic model for risk stratification in BC based on immunologic and hallmark genes sets, which presents a novel perspective on rational treatment approaches and accurate prognostic evaluations for BC by considering both tumor and adjacent non-tumor tissues. This highlights the importance of focusing on alterations in both tumor and adjacent non-tumor tissues, rather than solely on the tumor itself.