Clinical usefulness and factors affecting transcutaneous CO2 (PtcCO2) monitoring were studied in 22 critically ill neonates. In 10 of 22 infants, both tissue pH (pHt) and PtcCO2 were monitored simultaneously using Roche pHt and CO2 electrodes. Arterial blood gases were obtained from umbilical artery catheter. The effect of variations in blood pressure, oxygenation, arterial pH (pHa) pHt, and medications on PtcCO2 were studied. There was a linear correlation between pHt and pHa (n = 108, r = 0.85, slope 1.08), PtcCo2 and PaCO2 (n = 188, r = 0.85, slope 1.49). The pHt changes correlated better with PtcCO2 than with arterial CO2 (r = 0.78, slope 1.35). The major factors affecting the PtcCO2 relationship were (a) hypoxia and (b) acidosis. When PaO2 less than 40 torr, PtcCO2 correlated poorly to PaCO2 (n = 23, r = 0.48, slope 0.72) whereas, during normoxia (PaO2 greater than 40 torr), it correlated well, r = 0.85, slope 1.51 similarly. When pHt was greater than 7.30, PtcCO2 correlated better with PaCO2 (r = 0.88, slope 1.37) than during tissue acidosis (r = 0.71, slope 1.51). Mean blood pressure down to 30 mm Hg and administration of dopamine and tolazoline did not affect the PtcCO2 to PaCO2 relationship (n = 69, r = 0.86, slope 1.6). Clinically, both continuous pHt and PtcCO2 were found to be very useful in the management of critically ill neonates.