The historical method for the detection of botulinum neurotoxin (BoNT) is represented by the mouse bioassay (MBA) measuring the animal survival rate. Since the endpoint of the MBA is the death of the mice due to paralysis of the respiratory muscle, an ex vivo animal replacement method, called mouse phrenic nerve (MPN) assay, employs the isolated N. phrenicus-hemidiaphragm tissue. Here, BoNT causes a dose-dependent characteristic decrease of the contraction amplitude of the indirectly stimulated muscle. Within the EQuATox BoNT proficiency 13 test samples were analysed using the MPN assay by serial dilution to a bath concentration resulting in a paralysis time within the range of calibration curves generated with BoNT/A, B and E standards, respectively. For serotype identification the diluted samples were pre-incubated with polyclonal anti-BoNT/A, B or E antitoxin or a combination of each. All 13 samples were qualitatively correctly identified thereby delivering superior results compared to single in vitro methods like LFA, ELISA and LC-MS/MS. Having characterized the BoNT serotype, the final bath concentrations were calculated using the calibration curves and then multiplied by the respective dilution factor to obtain the sample concentration. Depending on the source of the BoNT standards used, the quantitation of ten BoNT/A containing samples delivered a mean z-score of 7 and of three BoNT/B or BoNT/E containing samples z-scores <2, respectively.