<p>Antibiotics are among the transformative advances in medicine, but many interact with mammalian cellular targets and pathways beyond their antimicrobial activity. A clinically important expression of these off-target effects is hematologic toxicity, including immune-mediated cytopenias and direct bone marrow suppression. This narrative review examines whether the same biology that injures normal hematopoietic cells can, in selected contexts, reveal therapeutically exploitable vulnerabilities in leukemia. We synthesize molecular, clinical, and preclinical evidence and organize it into an integrative framework linking mitochondrial translation inhibition, mitonuclear imbalance, oxidative phosphorylation failure, reactive oxygen species generation, DNA/topoisomerase stress, autophagy and lysosomal-flux blockade, and apoptosis modulation with both hematotoxicity and antileukemic activity. The strongest preclinical evidence supports selected tetracyclines, macrolides, and oxazolidinones, whereas evidence for beta-lactams, glycopeptides, polymyxins, rifamycins, fluoroquinolones, and folate-pathway agents remains more limited or largely hypothesis-generating. Importantly, antibiotic-induced cytopenia should not be interpreted as proof of leukemia selectivity: immune-mediated toxicity, supratherapeutic in vitro exposure, normal progenitor injury, pharmacokinetic constraints, microbiome effects, and resistance mechanisms all narrow the translational window. Overall, antibiotic hematotoxicity is best viewed as a biologically informative signal that can guide mechanism-based repurposing and combination strategies, but clinical development requires rigorous pharmacokinetic/pharmacodynamic validation, normal hematopoietic comparators, and biomarker-driven patient selection.</p>