Foodborne botulism is caused by the ingestion of foods containing botulinum neurotoxins (BoNTs). To study the heat stability of Clostridium botulinum neurotoxins, we needed to measure and compare the activity of botulinum neurotoxins, serotypes A and B, under various pasteurization conditions. Currently, the only accepted assay to detect active C. botulinum neurotoxin is an in vivo mouse bioassay, which raises ethical concerns with regard to the use of experimental animals. In this study, noninvasive methods were used to simultaneously detect and distinguish between active BoNT serotypes A and B in one reaction and sample. We developed an enzymatic activity assay employing internally quenched fluorogenic peptides corresponding to SNAP-25, for BoNT-A, and VAMP2, for BoNT-B, as an alternative method to the mouse bioassay. Because each peptide is labeled with different fluorophores, we were able to distinguish between these two toxins. We used this method to analyze the heat stability of BoNT-A and BoNT-B. This study reports that conventional milk pasteurization (63 °C, 30 min) inactivated BoNT serotype A; however, serotype B is heat-stable in milk and not inactivated by pasteurization. Using this activity assay, we also showed that the commonly used food processes such as acidity and pasteurization, which are known to inhibit C. botulinum growth and toxin production, are more effective in inactivating BoNT serotype A than serotype B when conventional pasteurization (63 °C, 30 min) is used.