Botulinum neurotoxins (BoNTs) are highly toxic proteases produced by anaerobic bacteria. Traditionally, a mouse bioassay (MBA) has been used for detection of BoNTs, but for a long time, laboratories have worked with alternative methods for their detection. One of the most promising in vitro methods is a combination of an enzymatic and mass spectrometric assay called Endopep-MS. However, no comprehensive validation of the method has been presented. The main purpose of this work was to perform a validation for the qualitative analysis of BoNT-A, B, C, C/D, D, D/C, and F in serum. The limit of detection (LOD), selectivity, precision, stability in matrix and solution, and correlation with the MBA were evaluated. The LOD was equal to or even better than that of the MBA for BoNT-A, B, D/C, E, and F. Furthermore, Endopep-MS was for the first time successfully used to differentiate between BoNT-C and D and their mosaics C/D and D/C by different combinations of antibodies and target peptides. In addition, sequential antibody capture was presented as a new way to multiplex the method when only a small sample volume is available. In the comparison with the MBA, all the samples analyzed were positive for BoNT-C/D with both methods. These results indicate that the Endopep-MS method is a valid alternative to the MBA as the gold standard for BoNT detection based on its sensitivity, selectivity, and speed and that it does not require experimental animals.