Objective:To investigate the value of short-time transcutaneous carbon dioxide pressure (TcPCO 2) and transcutaneous oxygen pressure (TcPO 2) monitoring in critically ill preterm infants. Methods:From January to December 2018, 62 critically ill neonates receiving respiratory support at Guangzhou Women and Children's Medical Center were retrospectively enrolled. A total of 348 sets of paired data including TcPCO 2/TcPO 2 and arterial carbon dioxide pressure (PaCO 2)/arterial oxygen partial pressure (PaO 2) were analyzed. The patients were divided into different groups based upon birth weight (23 cases>1 000 g-≤1 500 g, 129 sets of paired data; 18 cases≤1 000 g, 130 sets of paired data) and gestational age (16 cases born at ≤28 gestational weeks, 127 sets of paired data; 29 cases born at 28-34 gestational weeks, 159 sets of paired data) and the differences between groups were compared. The correlation and consistency of TcPCO 2/TcPO 2 and PaCO 2/PaO 2 were evaluated using Pearson's correlation and Blan-Altman scatter plots. Receiver operating characteristic (ROC) curve was drawn to analyze the diagnostic efficacy of TcPCO 2 in neonates with hypercapnia. Results:There was a positive correlation between TcPCO 2 and PaCO 2 in all patients ( r=0.913, 95% CI:0.894-0.929, P<0.05). In patients whose birth weight was>1 000 g-≤1 500 g or≤1 000 g, TcPCO 2 and PaCO 2 were positively correlated and the consistency were good ( r=0.909, 95% CI:0.874-0.935; r=0.934, 95% CI:0.908-0.953; both P<0.05), and the same finding was also observed in patients born at≤28 gestational weeks or 28-34 weeks of gestation ( r=0.938, 95% CI:0.913-0.956; r=0.871, 95% CI: 0.827-0.904; both P<0.05). The sensitivity, specificity and area under curve of TcPCO 2 in the diagnosis of hypercapnia were 90.91%, 85.85%, and 0.942, respectively. There was a poor correlation between TcPO 2 and PaO 2 in all patients and those with birth weight >1 000 g-≤1 500 g or gestational age 28-34 weeks (all r<0.75, all P<0.05). There was no correlation between TcPO 2 and PaO 2 in the birth weight ≤1 000 g and gestational age ≤28 weeks groups (both P>0.05). Conclusions:Short-time TcPCO 2 monitoring can accurately assess PaCO 2 in critically ill neonates requiring respiratory support and is of high diagnostic value for hypercapnia. However, TcPO 2 has limitation in evaluating PaO 2 and other indicators may need to be involved.