In 40 patients with esophageal varices, esophageal variceal pressure was assessed endoscopically using a pneumatic pressure sensor. The effects of vasopressin or nitroglycerin on variceal pressure and endoscopic findings were also assessed in two groups of seven patients. The results were as follows: (1) Variceal pressure was increased above 250 mmH2O in all patients who had bled, and the mean variceal pressure was significantly higher in patients who had bled than in those who had not (301 +/- 47 vs. 230 +/- 58 mmH2O respectively, p < 0.001). (2) Variceal pressure correlated with endoscopic findings, determined using the criteria of the Japanese Research Society for Portal Hypertension. It was significantly higher when varices with a feature of F2-F3 or RC(+2)-RC(+3) were compared to those with a feature of F1 or RC(-)-RC(+), respectively. (3) Both groups given vasopressin or nitroglycerin had significant reductions in variceal pressure; however, there was little improvement in endoscopic findings in those given nitroglycerin, compared to the improvement in those given vasopressin. Thus, use of a pneumatic pressure sensor proved to be a pertinent tool for assessing esophageal varices, along with endoscopic signs.