Variceal bleeding is the last step of a chain of events initiatedby an increase in portal pressure, followed by the development and progressive dilation ofvarices until these finally rupture and bleed. The ideal method to diagnose portal hypertension should be accurate, noninvasive, objective, and reproducible. The study evaluated the predictive value of two non-invasive parameters for the diagnosis of esophageal varices (EV): 1- Right liver lobe diameter/serum albumin ratios (RLLD /S. albumin), and 2- Platelet count/splenic bipolar diameter ratios (Platelets count/SBPD). This study included eighty Egyptian patients with post-hepatitic cirrhosis (45males and 35 females). They underwent laboratory ultrasono-graphic and endoscopic examinations within one week. RLLD/S. albumin and Platelets count/SBPD ratios were calculated. The results showed that EV were not detected by upper digestive endoscopy in 25%, while grade I of EV was
found in 17.5%, grade Ⅱ in 17.5%, grade Ⅲ in 20%, & grade Ⅳ in 20%. RLLD/S. albumin concentration ratio diagnosed the varices at cut off value of 3.43 with 95% sensitivity and 80% specificity. Also, it was positively correlated with grading of E.V, when this ratio increased the grading of E.V increases and vice versa. Besides, it predicted bleeding from E.V. at cut off value of 5.096 with 63% sensitivity and 73% specificity. Platelet count/ SBPD ratio predicted the presence of varices at cut off value 1847 with 95% sensitivity and 93% specificity, and negatively correlated with grading of EV, when this ratio decreased grading of E.V increase and vice versa. It also predicted bleeding from E.V. at cut off value of 4809 with 50% sensitivity and 93% specificity.