Background: VCO2 measurement is unreliable if an arterial-alveolar gradient (aADC2) exists We investigated the effect of several respiratory variables on aADCO2 in ventilated newborns .Subjects: 24 ventilated newborns (gestational age 24-37 ws, birth weight 450-2940 g). Interventions: End tidal CO2 was measured with a mainstream capnometer (CO2 SMO 7100, Novametrix) Arterial pCO2 was measured from indwelling catheters with Radiometer ABL610 gas analyzer. From these measures aADCO2 was calculated Associations between respiratory severity indexes and aADCO2 were assessed with regression analysis Results: aADCO2 increased with increasing oxigenation index (OI, r=0.551), alveolar-arterial difference (AaDO2, r=0.554) and mean airways preassure (MAP, r=0.324), and decreasing O2 arterial-alveolar ratio (aApO2, r=0.414). High aADCO2 (≥5 torr) was observed for OI≥3, MAP≥4 cmH2O, aApO2≥0.6, AaDO2≥50 torr. Conclusions: In ventilated newborns values of OI, MAP, and AaDO2 above (below for aApO2) the calculated cutoff values would make the estimates of VCO2 unreliable.