The main aetiology of acute gastrointestinal haemorrhage in cirrhotic patients is variceal bleeding. Prognosis is determined by early or late rebleeding rates and the severity of underlying liver disease, mostly estimated by Child Pugh score. Diagnosis and therapy of variceal bleeding is currently based on endoscopic sclerotherapy and more recently on banding ligation. However, the management of acute variceal bleeding remains controversial and vasoactive drugs are an alternative treatment. At present, most of these studies showed encouraging but conflicting results. These trials show that the cyclic octapeptide analogue (octreotide) of somatostatin or the triglycyl analogue (terlipressin) of vasopressin are safer and more effective than their natural drugs respectively. Clinically, drug choice depends on four factors: -results of trials comparing vasoactive treatment to classical sclerotherapy: comparison of these two kinds of treatment show similar results concerning haemostatic rate as well as mortality especially for somatostatin or its synthetic analogue; -results of trials comparing synthetic analogue of vasopressin to cyclic analogue of somatostatin in variceal bleeding: current study designs demonstrate an arithmetic difference (p = 0.06) with a better early haemostatic rate after octreotide associated with comparable final haemostasis (after 24 hours) and mortality; -results of combination of both treatments (i.e. sclerotherapy associated with vasoactive drug versus sclerotherapy): such association decreases variceal rebleeding without improvement of survival rate; -and finally, importance of adverse drug effects on hepatic and renal functions: few studies show scanty and conflicting adverse drug effect especially on free water clearance which must be studied by further clinical trials to confirm their benefit in emergency management of variceal bleeding.