BACKGROUND Due to its bacteriological spectrum and efficacy in skin and soft tissue infections, ceftobiprole may be of interest for extracorporeal membrane oxygenation (ECMO) cannula-related infection. It is unknown whether ceftobiprole pharmacokinetics (PK) are changed by ECMO. METHODS A retrospective monocentric cohort study was performed of 35 patients with suspected ECMO-related cannula infections (28 on ECMO, seven after ECMO removal), who received ceftobiprole as empiric treatment and had ceftobiprole blood levels measured at trough, peak and CT50 (50% of the dosing interval). Ceftobiprole blood levels of the 28 patients on ECMO were compared with those of the seven patients without ECMO. Factors associated with low ceftobiprole trough levels were also explored. RESULTS Among the 35 patients included, 29 had a confirmed cannula-related infection and 48 pathogens were isolated. Ceftobiprole MIC was determined in 29 of these 48, and 23 (79%) were susceptible to ceftobiprole. Ceftobiprole blood levels (at trough, peak and CT50) were similar in ECMO and non-ECMO patients. Moreover, in patients whose pathogens responsible for infection were susceptible to ceftobiprole, 94% had a ceftobiprole trough level above the MIC. Ceftobiprole blood levels were decreased in patients with acute renal failure requiring renal replacement therapy (RRT) and in those with increased renal clearance (defined as creatinine clearance > 130 mL/min), independent of ECMO. No other factor was associated with modification of ceftobiprole PK/pharmacodynamics (PK/PD). CONCLUSIONS The ceftobiprole PK/PD was no different in patients during ECMO or after its withdrawal. Factors associated with decreased ceftobiprole blood levels were patients requiring RRT and those with increased renal clearance.