The objective of this article is to illustrate both the potential and the limitations of molecular imaging in stroke research. By molecular imaging we mean the visual representation of biological processes at the cellular and molecular level. The use of molecular imaging for stroke diagnosis is still at a very preliminary stage and many of these procedures have only been tested in animals. In rats, stroke therapy using stem cells can be monitored by magnetic resonance imaging (MRI), green fluorescent protein (GFP) or luciferase (LUC) imaging. The migration of macrophages, which take up intravenously administered iron-based contrast agents and then migrate to the area of infarction, can already be observed in stroke patients. With MRI, the new agent Gd-DTPA-sLexA that binds to E- and Pselectin can specifically visualize selectin-mediated early endothelial activation after transient focal ischemia “in vivo”. Decreased glial fibrillary acidic protein (GFAP) gene expression can be imaged in vivo by scintigraphy 24 hours after cerebral ischemia using a peptide nucleic acid antisense conjugate labeled with 111In and that hybridizes to the rat GFAP mRNA. Technetium-99m hydrazine nicotinamide- labeled HYNIC-annexin V SPECT can not only detect sites of neuronal injury in stroke patients but also can monitor the effects of neuroprotective therapy with a monoclonal antibody raised against FasLigand (FasL) in rats. Finally, information about cell metabolism in the infarct region can be gained using certain intracellular tracers [e.g. 18F-fluoromisonidazole (FMISO)]. Imaging benzodiazepine receptors with 11C-flumazenil (FMZ) can distinguish between irreversibly damaged and viable penumbra tissue early after stroke.