Résumé : Antimicrobial resistance (AMR) is a worldwide healthcare emergency. Whether insufficient beta-lactam antibiotic concentrations can be associated with AMR emergence remains controversial. This is a retrospective single-center cohort study including patients admitted to the intensive care unit of a tertiary university hospital from 2009 to 2014, who required a broad-spectrum beta-lactam antibiotic and had at least one therapeutic drug monitoring (TDM). Patients were categorized as having inadequate drug levels if the trough concentration (Cmin) fell below the clinical breakpoint for Pseudomonas aeruginosa. AMR was defined according to breakpoints recommended by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) using the disk diffusion method. A total of 444 patients (male sex, n = 313, 71%; female sex, n = 131, 29%; mean age 58 ± 15 years) were enrolled in the study. Patients received piperacillin/tazobactam (n = 168), ceftazidime/cefepime (n = 58) or meropenem (n = 218); among them, 65 (15%) had insufficient drug levels. Nine of these 65 (13.8%) patients with insufficient antibiotic levels acquired at least one pathogen with AMR within 15 days of TDM, when compared to 84/379 (22%) in the other group (OR 0.56 [95%CI 0.27–1.19]; p = 0.13). In a multivariable competing-risk analysis including male gender, APACHE score on admission, previous colonization by other MDR bacteria, urinary catheter, central venous catheter, mechanical ventilation, previous hospitalization and previous surgery, insufficient antibiotic levels were not associated with AMR acquisition (sHR 0.84 [95% CI 0.42–1.68]). Similar results were found when a higher threshold was used to define insufficient drug levels (Cmin < 4 times the clinical breakpoint). In conclusion, insufficient beta-lactam levels were not independently associated with AMR acquisition. Future prospective studies are needed to evaluate better the relationship between low drug levels and antibiotic resistance acquisition.