Introduction: Pressure support ventilation (PSV) should allow spontaneous breathing with a “normal” neuro-ventilatory drive. Low neuro-ventilatory drive puts the patient at risk of diaphragmatic atrophy while high neuro-ventilatory drive may causes dyspnea and patient self-inflicted lung injury. We continuously assessed for 12h the electrical activity of the diaphragm (EAdi), a close surrogate of neuro-ventilatory drive, during PSV. Our aim was to document the EAdi trend and the occurrence of periods of “Low” and/or “High” neuro-ventilatory drive during clinical application of PSV. Method: In 16 critically ill patients ventilated in the PSV mode for clinical reasons, inspiratory peak EAdi peak (EAdiPEAK), pressure time product of the trans-diaphragmatic pressure per breath and per minute (PTPDI/b and PTPDI/min, respectively), breathing pattern and major asynchronies were continuously monitored for 12h (from 8 a.m. to 8 p.m.). We identified breaths with “Normal” (EAdiPEAK 5–15μV), “Low” (EAdiPEAK < 5μV) and “High” (EAdiPEAK > 15μV) neuro-ventilatory drive. Results: Within all the analyzed breaths (177.117), the neuro-ventilatory drive, as expressed by the EAdiPEAK, was “Low” in 50.116 breath (28%), “Normal” in 88.419 breaths (50%) and “High” in 38.582 breaths (22%). The average times spent in “Low”, “Normal” and “High” class were 1.37, 3.67 and 0.55h, respectively (p < 0.0001), with wide variations among patients. Eleven patients remained in the “Low” neuro-ventilatory drive class for more than 1h, median 6.1 [3.9–8.5]h and 6 in the “High” neuro-ventilatory drive class, median 3.4 [2.2–7.8]h. The asynchrony index was significantly higher in the “Low” neuro-ventilatory class, mainly because of a higher number of missed efforts. Conclusions: We observed wide variations in EAdi amplitude and unevenly distributed “Low” and “High” neuro ventilatory drive periods during 12h of PSV in critically ill patients. Further studies are needed to assess the possible clinical implications of our physiological findings.