Immunotherapy has revolutionized cancer treatment; however, predicting patient response remains a significant challenge. Our study identified a novel plasma cell signature, Plasma cell.Sig, through a pan-cancer single-cell RNA sequencing analysis, which predicts patient outcomes to immunotherapy with remarkable accuracy. The signature was developed using rigorous machine learning algorithms and validated across multiple cohorts, demonstrating superior predictive power with an area under the curve (AUC) exceeding 0.7. Notably, the low-risk group, as classified by Plasma cell.Sig, exhibited enriched immune cell infiltration and heightened tumor immunogenicity, indicating an enhanced responsiveness to immunotherapy. Conversely, the high-risk group showed reduced immune activity and potential mechanisms of immune evasion. These findings not only enhance understanding of the intrinsic and extrinsic immune landscapes within the tumor microenvironment but also pave the way for more precise, biomarker-guided immunotherapy approaches in oncology.