Commercial rabies vaccines, used by veterinarians in the Netherlands, were collected for testing in the mouse potency test. Of the six vaccines tested, two were clearly below the minimal requirements for potency of 1.0 IU. Of these six vaccines the rabies virus glycoprotein (GP) and nucleoprotein (NP) contents were determined in an antigen competition ELISA. The GP content proved to correlate well with the potency found in the mouse potency test (r = 0.95, p < 0.01), whereas no such correlation was found for the NP content (r approximately 0, p > 0.05). After the manufacturers were told about the results, one of the two vaccines that did not comply with the requirements was withdrawn from the market. Measurement of the GP content of a second lot of the remaining vaccines indicated that sufficiently high levels of GP were present in all five. Additional in vivo testing in mice for efficacy against intracerebral challenge with the Dutch bat rabies virus EBL1-12 resulted in acceptable levels of protection with four of these five vaccines of the second lot. The data presented illustrate the need for continued potency evaluation of veterinary rabies vaccines in the Netherlands.