Background: Intrahepatic cholangiocarcinoma (iCCA) is subclassified by the presence or absence of potentially actionable molecular alterations, of which IDH1 mutations and FGFR2 fusions and rearrangements are most frequently observed. Little is known about the relationship between molecular features of iCCAs and the tumor immune microenvironment (TME). Methods: We performed a high-parameter immune phenotyping of the TME of FGFR2 fusion/rearrangement positive (FGFR2+), IDH1 mutation positive (IDH1+), and FGFR2/IDH1 wild type iCCA samples at the single-cell level using CO-Detection by indEXing (CODEX). Neighborhood analysis was based on a k-nearest-neighbors approach followed by k-means clustering to identify the spatial location of each cell type and their proximity to one another. Simplified spatial networks specific to molecular subtypes were generated based on the shortest Euclidean distances from any given cell to all other cell types. Results: A total of 24 tumors were examined (n=7 FGFR2+, n=9 IDH1+, n=8 other). Compared to other molecular subtypes, FGFR2+ tumors had an immune-desert phenotype characterized by a reduced abundance of CD8+ T cells (p < 0.05), CD117+ mast cells (p < 0.05), and CD68+/CD163+ macrophages (p < 0.05) but with an increased number of CD11b+/CD15+ granulocytes (p < 0.05). IDH1+ tumors were characterized by a decreased amount of CD11b+/CD15+ granulocytes (p < 0.05) when compared to other subtypes. In terms of spatial analysis, CD4+ T cells were located further from tumor cells in FGFR2+ iCCA compared to IDH1+ iCCA (p < 0.05) and other molecular subtypes (p=0.05). In contrast, CD11b+/CD15+ granulocytes were located closer to tumor cells in FGFR2+ tumors (p < 0.05). Additionally, closer neighbor proximity between CD4+ T cells and CD11b+/CD15+ granulocytes was observed in FGFR2+ tumors compared to both IDH1+ tumors (p < 0.05) and other tumor subtypes (p = 0.05), indicating a stronger lymphoid–myeloid neighboring relationship in the TME of FGFR2+ iCCA. Conclusions: FGFR2+ and IDH1+ iCCAs have distinct immunophenotypes, providing initial evidence that molecular alterations may inform the selection of patients for immunotherapy. Both the quantity of immune cells as well as the spatial relationships between immune cells may influence iCCA prognosis and response to therapy. Tailoring immunotherapeutic approaches to specific molecular subsets may improve outcomes across the divergent molecularly defined iCCA subtypes.