BACKGROUND: Local anesthetics promote anticancer immune responses. A machine learning-based algorithm trained with information on the biological effects and molecular descriptors of analgesics, anesthetics, hypnotics and opioids predicted antitumor effects for dexmedetomidine (DEX). DEX is a sedative acting as an alpha2-adrenoceptor (ADRA2) agonist. Based on these premises, we investigated the putative antineoplastic effects of DEX. RESULTS: In vitro, DEX promoted premortem stresses such as autophagy and partial endoplasmic reticulum stress with the phosphorylation of eukaryotic initiation factor 2 alpha and the inhibition of the splicing of X-box binding protein 1. DEX elicited the biomarkers of immunogenic cell death, including the release of ATP and high-mobility group box 1 protein, and the cell surface exposure of calreticulin, enhancing the engulfment of malignant cells by dendritic cells. In immunocompetent mice, DEX decreased the progression of colorectal cancers, fibrosarcomas, mammary carcinomas and melanomas, as it improved overall survival. These effects were inhibited by the ADRA2 antagonist yohimbine, suggesting that DEX mediates its anticancer effects at least in part on-target. Depending on the specific tumor model, DEX also enhanced the cytotoxic T cell/regulatory T cell ratio in the tumor bed and draining lymph nodes. Programmed cell death protein 1 blockade tended to improve DEX effects. After rechallenge with antigenically identical cells, no tumor appeared, indicating the formation of immunological memory. CONCLUSIONS: These results confirm the machine learning-predicted anticancer activity of DEX. Beyond its utility as a sedative agent in oncological intensive care, DEX may improve anticancer immunosurveillance and sensitize tumors to immune checkpoint blockade.