Ｏｂｊｅｃｔｉｖｅ　To compare the difference between TL-300 continuous non-invasive arterial pressure (CNAP) and invasive arterial pressure (IAP) monitoring.　Methods　Seventy and two patients undergoing surgery under general anesthesia were enrolled at our study. Invasive radial pressure monitoring was established in the non-dominant hand. CNAP monitoring (TL-300) used a pressure sensor directly over the radial artery at the wrist before induction of anesthesia. The Bland-Altman plot was used to determine the bias of blood pressure measurements between CNAP with TL-300 and IAP during general anesthesia maintenance.　Results　We obtained 433 valid pressure readings from 72 patients. For SBP, the bias (SD of bias) was (-0.6±7.0) mmHg (limits of agreement: -14.6-13.4 mmHg, 1 mmHg=0.133 kPa). For DBP, the bias (SD of bias) was (-7.8±9.8) mmHg (limits of agreement: -27.4 -11.8 mmHg). For MAP, the bias (SD of bias) was (-5.8±6.2) mmHg (limits of agreement: -18.2-6.6 mmHg). The percentage within limits of agreement was 95.2%, 96.1%, 94.9%, respectively. Pearson correlation analysis was made for SBP, DBP, and MAP. Invasive arterial pressures measurement were closely correlated with continuous non-invasive arterial pressures measurement. The correlation coefficient was 0.927, 0.711, 0.903, respectively (P<0.01).　Conclusions　High correlation and good agreement of blood pressure measurement exists between CNAP with TL-300 and IAP monitoring during general anesthesia.