BACKGROUND: IMbrave150 has established the superiority of atezolizumab plus bevacizumab over sorafenib in patients with unresectable hepatocellular carcinoma (HCC). METHODS: We generated a prospectively maintained database including patients treated with atezolizumab plus bevacizumab for unresectable HCC across Europe, Asia and USA. Clinico-pathologic characteristics were assessed for their prognostic influence on overall survival (OS) and progression-free survival (PFS) in univariable and multivariate analyses. Overall response rate by RECIST v1.1 and treatment-related adverse events (TRAEs) per CTCAE v.5.0 were reported. RESULTS: Out of 433 patients, 296 Child-Pugh A and ECOG performance status01 patients received atezolizumab plus bevacizumab in first line and were included. Patients were mostly male (82.7%), cirrhotic (75%) with history of viral hepatitis (65.9%). Overall, 68.9% had Barcelona Clinic Liver Cancer C-stage HCC with portal vein tumour thrombosis (PVTT, 35%) and extrahepatic spread (EHS, 51.7%). After a median follow-up of 10.0 months (95% confidence interval (CI): 9.4-10.4), median OS and PFS were 15.7 (95% CI: 14.5-NE) and 6.9 months (95% CI: 6.1-8.3), respectively. In the response-evaluable patients (n = 273), overall response rate was 30.8%. Overall, 221 patients (74.6%) developed TRAEs, with 70 (23.6%) reporting grade 3 or higher TRAEs; 25 (8.4%) patients had bleeding events. OS was independently associated with baseline Albumin-bilirubin (ALBI) grade and PVTT. Shorter PFS was associated with AFP≥ 400 ng/ml, worse ALBI and presence of EHS. CONCLUSION: This global observational study confirms the reproducible safety and efficacy of atezolizumab plus bevacizumab in routine clinical practice. Within Child-Pugh-A criteria, the presence of PVTT and higher ALBI grade identify patients with poorer survival.