The hepatic venous pressure gradient (HVPG) clearly reflects portal pressure in cirrhotic portal hypertension. Its relation with variceal bleeding has been well studied. We undertook to study the relation of HVPG to variceal size, Child's status, and etiology of cirrhosis. Patients with cirrhotic portal hypertension with esophageal varices underwent HVPG measurement as part of a prospective evaluation. One hundred seventy-six cirrhotics with varices (M:F, 140:36; mean age, 42.6 +/- 13.4 years), 104 with CLD related to viral etiology, 40 with alcoholic liver disease, 26 cryptogenic with cirrhosis, and 6 with miscellaneous causes of CLD underwent HVPG measurement. The mean HVPG was lower in patients with small varices (n = 77; 14.6 +/- 5.9 mm Hg) than in patients with large varices (n = 99; 19.2 +/- 6.6 mm Hg; P < 0.01). In patients with large varices, the mean HVPG in bleeders (n = 37) was higher than in nonbleeders (n = 62) (21.7 +/- 7.2 vs 17.9 +/- 6.2 mm Hg; P < 0.01). The mean HVPG was significantly higher in Child's B (n = 97; 17.4 +/- 6.9 mm Hg) and C (n = 56; 19.0 +/- 5.7 mm Hg) compared to Child's A cirrhotics (n = 23; 12.2 +/- 5.9 mm Hg; P < 0.01), and Child's C compared to Child's B cirrhotics (P = 0.05). HVPG was higher in alcoholic compared to nonalcoholic cirrhotics (20.8 +/- 7.3 vs 16.4 +/- 6.3 mm Hg; P < 0.05), but this was not significant in multivariate analysis. The HVPG was comparable between hepatitis B- and hepatitis C virus-related cirrhotics (P = 0.8). Cirrhotics with ascites had a higher HVPG than those without ascites (18.5 +/- 5.6 vs 16.6 +/- 7.6 mm Hg; P = 0.02). In multivariate analysis, only Child's status, size of varices, and variceal bleed predicted higher HVPG. HVPG is higher in cirrhotics with large varices and a history of bleed. There is a good correlation between HVPG and large varices, bleeder status, and ascites. A higher HVPG reflects more severe liver disease. The etiology of liver disease did not influence the portal pressure.