Background: Whether sepsis-associated diaphragm dysfunction may improve despite the exposure of mechanical ventilation in critically ill patients is unclear. This study aims at describing the diaphragm function time course of septic and non-septic mechanically ventilated patients.Methods Secondary analysis of two prospective observational studies of mechanically ventilated patients in whom diaphragm function was assessed twice: within the 24 h after intubation and when patients were switched to pressure support mode, by measuring the endotracheal pressure in response to bilateral anterior magnetic phrenic nerve stimulation (Ptr,stim). Change in diaphragm function was expressed as the difference between Ptr,stim measured under pressure support mode and Ptr,stim measured within the 24 h after intubation. Sepsis was defined according to the Sepsis-3 international guidelines upon inclusion. In a sub-group of patients, the right hemidiaphragm thickness was measured by ultrasound.Results Ninety-two patients were enrolled in the study. Sepsis upon intubation was present in 51 (55%) patients. In septic patients, primary reason for ventilation was acute respiratory failure related to pneumonia (37/51; 73%). In non-septic patients, main reasons for ventilation were acute respiratory failure not related to pneumonia (16/41; 39%), coma (13/41; 32%) and cardiac arrest (6/41; 15%). Ptr,stim within 24 h after intubation was lower in septic patients as compared to non-septic patients: 6.3 (4.9–8.7) cmH 2 O vs. 9.8 (7.0–14.2) cmH 2 O ( p = 0.004), respectively. The median (interquartile) duration of mechanical ventilation between first and second diaphragm evaluation was 4 (2–6) days in septic patients and 3 (2–4) days in non-septic patients ( p = 0.073). Between first and second measurements, the change in Ptr,stim was + 19% (− 13–61) in septic patients and − 7% (− 40–12) in non-septic patients ( p = 0.005). In the sub-group of patients with ultrasound measurements, end-expiratory diaphragm thickness decreased in both, septic and non-septic patients. The 28-day mortality was higher in patients with decrease or no change in diaphragm function.Conclusion Septic patients were associated with a more severe but reversible impaired diaphragm function as compared to non-septic patients. Increase in diaphragm function was associated with a better survival.