It has been reported that distal cerebrospinal fluid-contacting neurons (dCSF-CNs) can be detected by immunohistochemical assay using cholera toxin subunit B-conjugated horseradish peroxidase (CB-HRP). In the present study, another two methods with CB alone or CB-conjugated FITC (CB-FITC) were used, and the results from the three methods were compared. Adult Sprague-Dawley rats were randomly divided into three groups with CB-HRP, CB or CB-FITC. Tracers were diluted to 30% in artificial cerebrospinal fluid and injected separately (in a volume of 3 L) into the lateral ventricle. Animals from the CB-HRP and CB groups were perfused 48 h after surgery while animals from the CB-FITC group were perfused 1, 3, 6, 12, 24 or 48 h after surgery. The brain was sectioned (40 m) for immunofluorescence and five sections with positive neurons were selected from each rat for neuron counting. Three clusters of positive neurons in a 'Y-like' distribution were detected ventral to the cerebral aqueduct of rats from the three groups. No significant difference was observed among the quantitative data. In the CB-FITC group, stable staining was detected even at 6 h after injection. Taken together, lateral ventricle injection of CB/CB-FITC is a useful method for labeling dCSF-CNs in rats. The CB-FITC method makes dCSF-CNs labeling much simpler and more convenient.