Treatment allocation in patients with hepatocellular carcinoma (HCC) on a background of Child-Pugh B (CP-B) cirrhosis is controversial. Liver resection has been proposed in small series with acceptable outcomes, but data are limited. The aim of this study was to evaluate the outcomes of patients undergoing liver resection for HCC in CP-B cirrhosis, focusing on the surgical risks and survival. Patients were retrospectively pooled from 14 international referral centers from 2002 to 2017. Postoperative and oncological outcomes were investigated. Prediction models for surgical risks, disease-free survival and overall survival were constructed. A total of 253 patients were included, of whom 57.3% of patients had a preoperative platelet count <100,000/mm3, 43.5% had preoperative ascites, and 56.9% had portal hypertension. A minor hepatectomy was most commonly performed (84.6%) and 122 (48.2%) were operated on by minimally invasive surgery (MIS). Ninety-day mortality was 4.3% with 6 patients (2.3%) dying from liver failure. One hundred and eight patients (42.7%) experienced complications, of which the most common was ascites (37.5%). Patients undergoing major hepatectomies had higher 90-day mortality (10.3% vs. 3.3%; p = 0.04) and morbidity rates (69.2% vs. 37.9%; p <0.001). Patients undergoing an open hepatectomy had higher morbidity (52.7% vs. 31.9%; p = 0.001) than those undergoing MIS. A prediction model for surgical risk was constructed (https://childb.shinyapps.io/morbidity/). The 5-year overall survival rate was 47%, and 56.9% of patients experienced recurrence. Prediction models for overall survival (https://childb.shinyapps.io/survival/) and disease-free survival (https://childb.shinyapps.io/DFsurvival/) were constructed. Liver resection should be considered for patients with HCC and CP-B cirrhosis after careful selection according to patient characteristics, tumor pattern and liver function, while aiming to minimize surgical stress. An estimation of the surgical risk and survival advantage may be helpful in treatment allocation, eventually improving postoperative morbidity and achieving safe oncological outcomes. Liver resection for hepatocellular carcinoma in advanced cirrhosis (Child-Pugh B score) is associated with a high rate of postoperative complications. However, due to the limited therapeutic alternatives in this setting, recent studies have shown promising results after accurate patient selection. In our international multicenter study, we provide 3 clinical models to predict postoperative surgical risks and long-term survival following liver resection, with the aim of improving treatment allocation and eventually clinical outcomes. Copyright © 2019. Published by Elsevier B.V.