In hospitalized patients with advanced chronic liver disease (ACLD), we aimed to evaluate the association between myostatin and muscle mass, its relation to inflammation and to assess the added prognostic value of myostatin for survival. In a prospective single-center cohort study, inclusion criteria were: consecutive hospitalization for ACLD and Child-Pugh score ≥ 7 points. Baseline parameters were myostatin, C-reactive protein (CRP), hand-grip strength (HGS), mid-arm muscle circumference (MAC), transversal psoas muscle index (TPMI). Patients were followed-up for at least 400 days. We included 198 men, 157 women, and 40 controls, median myostatin levels in pg/ml were 1790.1 in women, 1959.4 in men, and 3850.0 and 2996.0 in healthy men and women. Myostatin positively correlated with TPMI, but weakly with MAC and HGS, and not in women. Myostatin negatively correlated with CRP in both genders. In cases with CRP 10 mg/l, regression analysis of myostatin versus HGS, MAC or TPMI showed steeper dependence curve. During follow up, 85 men and 64 women (42.9% versus 40.8%) have died, 22 men and 19 women (11.1% versus 12.1%) underwent liver transplantation. Cumulative incidence of death was higher in men with myostatin levels < 1600.0 pg/ml, but not in women. In men, MELD score and myostatin cut-off were independent predictors of worse survival but did not predict survival in women. In men, myostatin levels directly reflect the muscle mass and low levels independently predict worse survival. In women, myostatin is not associated with muscle mass or survival.