AIM/METHODS: Variceal pressure was measured at endoscopy in 98 patients with a non-invasive, pressure-sensitive gauge technique. RESULTS: In nine patients undergoing sclerotherapy, the values obtained correlated closely with those determined by fine-needle puncture of the varices (r = 0.95, p < 0.001). In 15 patients receiving placebo as part of a double-blind, placebo-controlled prospective evaluation of prophylactic treatment with propranolol, variceal pressure did not change significantly over a 1-year period: 16.5 +/- 3.2 mmHg at onset: 16.7 +/- 3.4 after 3 months and 15.9 +/- 3.5 after 12 months. In contrast, propranolol therapy given to 15 patients decreased the pressure from 16.2 +/- 3.0 mmHg before therapy to 13.5 +/- 2.0 after 3 months (p < 0.001) and to 13.1 +/- 4.1 after 12 months (p < 0.05). Patients with a recent bleeding episode had higher pressures than those who had not yet experienced any haemorrhage (19.3 +/- 3.5 mmHg, n = 17 vs 13.9 +/- 3.3, n = 62) (p < 0.001). This was confirmed in a prospective study: the last variceal pressure in bleeders and non-bleeders was respectively 17.2 +/- 1.7 mmHg (n = 4) and 14.8 +/- 4.1 mmHg (n = 26) (p < 0.01). Overall, variceal pressure measurements correlated with the size of varices, the presence of red colour signs and with the North Italian Endoscopic Club score. CONCLUSIONS: This study has shown that the non-invasive pressure gauge Varipress represents a reproducible method to regularly assess variceal pressure, an important parameter related to the risk of bleeding, and to assess the effects of pharmacological therapy with propranolol.