Ventilator-associated pneumonia is the leading nosocomial infection in the intensive care units, and is associated with prolonged mechanical ventilation and overuse of antibiotics. Initiating antibiotic therapy immediately after bacteriological sampling (immediate strategy) may expose uninfected patients to unnecessary treatment, while waiting for bacteriological confirmation (conservative strategy) may delay ventilator-associated pneumonia in infected patients. The decision to start antibiotic therapy for ventilator-associated pneumonia takes three points into account: diagnostic probability, the risks to the patient if Antibiotic Therapy is delayed, and the risk of selection of resistant bacteria. Diagnostic probability is limited, given the subjective and non-specific nature of the diagnostic criteria, and only 30-50% of suspected cases are confirmed bacteriologically (whereas samples are only taken when the pre-test probability is sufficient). The risks associated with delayed antibiotic therapy are unknown, as few observational studies have directly assessed the impact of the timing of Antibiotic Therapy initiation on outcome (frequent confusion between delayed and inappropriate Antibiotic Therapy). Iregui et al. found that delaying Antibiotic Therapy by more than 24 hours was associated with higher mortality. However, more recent before-and-after studies have shown that the conservative strategy was associated with lower mortality, more frequently appropriate initial Antibiotic Therapy and shorter duration of Antibiotic Therapy. Similarly, in a recent before-and-after study by our team, initiating antibiotic therapy only upon microbiological confirmation of ventilator-associated pneumonia without septic shock or severe acute respiratory distress syndrome was not associated with an increase in ventilation time, length of stay or excess mortality at D28; but was associated with antibiotic therapy that was more often appropriate (DELAVAP, MARTIN et al, Annals of Intensive Care, 2024). Finally, the recent multicenter TARPP pilot study in surgical intensive care suggests that antibiotic therapy initiated on the basis of microbiological data in patients with suspected ventilator-associated pneumonia not requiring vasopressor support is not associated with a poorer outcome than immediate antibiotic therapy without documentation (the only randomized study on this subject). Antibiotic Therapy for suspected ventilator-associated pneumonia that is not subsequently confirmed is an unnecessary use of antibiotics and carries a risk of selection of resistant bacteria, with adverse effects on public health. It has been reported that a conservative Antibiotic Therapy prescription strategy for intensive care units -acquired infections reduces Antibiotic Therapy use and the incidence of acquired β-lactamase-producing Enterobacteriaceae infections. Overall, in patients with suspected ventilator-associated pneumonia but no signs of clinical severity, given the uncertainty about attributable mortality and concerns about bacterial resistance, the evaluation of the conservative Antibiotic Therapy strategy is reasonable. Some French intensive care units already delay Antibiotic Therapy until confirmation of ventilator-associated pneumonia, except in patients with severe hypoxemia or the need for vasopressor support.