The increasing prevalence of microorganisms producing extended-spectrum β-lactamases (ESBLs) poses a critical global challenge in combating antibiotic resistance. Rapid and innovative approaches are essential for their detection and identification. Our objective was to optimise a phenotypic method for straightforward and rapid ESBL detection in strains obtained from food environments, employing MALDI-TOF mass spectrometry. A collection of over 150 enterobacterial isolates, including 60 ESBL producers, 60 sensitive strains, and 25 constitutive AmpC producers, was analysed in triplicate using the optimised method. Bacterial colonies were cultured on Mueller-Hinton Agar for 24 hours, mixed with cefotaxime, and incubated at 37 ºC during 60 min by shaking. The suspension was then centrifuged, and the supernatant was directly analysed on the MALDI-TOF MS target, together with an aliquot of HCCA matrix (Bruker). The method targeted two types of peaks: intact antibiotic peaks (456 and 478 m/z) and lysed antibiotic peaks (370 and 414 m/z). The ratio (log value) between the lysed and intact antibiotic peaks determined the strain's ability to degrade the antibiotic. The results robustly differentiated ESBL producers and sensitive isolates. Notably, constitutive AmpC isolates, often overlooked in rapid detection methods, exhibited overlap with both "strongest" sensitive profiles and "weakest" ESBL profiles. Therefore, we established a threshold requiring confirmation by an alternative phenotypic method for these isolates. In conclusion, our optimised method provides a quick and accurate approach to detect ESBL-producing enterobacteria isolated from food environments.