Diaphragmatic insults occurring during intensive care unit (ICU) stays have become the focus of intense research. However, diaphragmatic abnormalities at the initial phase of critical illness remain poorly documented in humans. To determine the incidence, risk factors, and prognostic impact of diaphragmatic impairment on ICU admission. Prospective, 6-month, observational cohort study in two ICUs. Mechanically ventilated patients were studied within 24 hours after intubation (Day 1) and 48 hours later (Day 3). Seventeen anesthetized intubated control anesthesia patients were also studied. The diaphragm was assessed by twitch tracheal pressure in response to bilateral anterior magnetic phrenic nerve stimulation (Ptr,stim). Eighty-five consecutive patients aged 62 (54-75) (median [interquartile range]) were evaluated (medical admission, 79%; Simplified Acute Physiology Score II, 54 [44-68]). On Day 1, Ptr,stim was 8.2 (5.9-12.3) cm H2O and 64% of patients had Ptr,stim less than 11 cm H2O. Independent predictors of low Ptr,stim were sepsis (linear regression coefficient, -3.74; standard error, 1.16; P = 0.002) and Simplified Acute Physiology Score II (linear regression coefficient, -0.07; standard error, 1.69; P = 0.03). Compared with nonsurvivors, ICU survivors had higher Ptr,stim (9.7 [6.3-13.8] vs. 7.3 [5.5-9.7] cm H2O; P = 0.004). This was also true for hospital survivors versus nonsurvivors (9.7 [6.3-13.5] vs. 7.8 [5.5-10.1] cm H2O; P = 0.004). Day 1 and Day 3 Ptr,stim were similar. A reduced capacity of the diaphragm to produce inspiratory pressure (diaphragm dysfunction) is frequent on ICU admission. It is associated with sepsis and disease severity, suggesting that it may represent another form of organ failure. It is associated with a poor prognosis. Clinical trial registered with www.clinicaltrials.gov (NCT 00786526).