Objective The study aimed to explore a specific marker to label cerebrospinal fluid-contacting neurons (CSF-CNs) in rat and provide a new way to identify CSF-CNs.Methods Adult Sprague-Dawley rats, male, were randomly received 4 different treatment (n=6): A.Cholera toxin subunit B (CB) injection into the cerebrospinal fluid-contacting nucleus (CSF-CN); B.CB injection into the hippocampus(non-CSF-CN) as control; C.CB injection into the lateral ventricle; D.Classical labeling method (lateral ventricle injection of CB-conjugated horseradish peroxidase (CB-HRP)) as a positive control.Rats received same volume of CB or CB-HRP (1 μL).Animals were perfused after 48 h survival and the relevant brain tissue was sectioned (40 μm) for immunofluorescence to observe the effects of different labeling methods.Results Group A: Neurons labeled by CB clustered only in the CSF-CN, not in other areas; Group B: In the hippocampus (non-CSF-CN area), there were few CB positive neurons; Group C: CB positive neurons clustered in the CSF-CN; D Group: Neurons labeled by CB-HRP appeared only in the CSF-CN.That is, the method of CB injected into lateral ventricle has the same effects on labeling area with CB-HRP; the method of CB injected into CSF-CN has the same effects with C/D group; the method of CB injected into non-CSF-CN showed that neither CSF-CNs nor non-CSF-CNs was labeled.Conclusion Parenchyma injection of CB is a scientific and reliable method to label CSF-CNs in rats.