AIM: Recent advances in next-generation sequencing (NGS) technologies allow for evaluation of genetic alterations in various cancer-related genes in daily clinical practice. Archival formalin-fixed paraffin-embedded (FFPE) tumor tissue is often used for NGS-based clinical sequencing assays; however, the success rate of NGS assays using archival FFPE tumor tissue is reported to be lower than that using fresh tumor tissue. We aimed to evaluate the feasibility and safety of ultrasound (US)-guided liver tumor biopsy for NGS-based multiplex gene assays. METHODS: We compared the success rate of NGS assays between archival FFPE tumor tissues and US-guided liver tumor biopsy tissues, and summarized the treatment progress of the patients. RESULTS: Next-generation sequencing assays using US-guided liver biopsy samples were successful in all patients (22/22), whereas the success rate with archival FFPE tumor tissue was 84.8% (151/178, P < 0.05). At least one potentially actionable genetic alteration was identified from the US-guided liver biopsy samples in 20 of 22 patients. Among the 18 patients with actionable genetic alterations targetable with drugs approved by the US Food and Drug Administration, eight initiated mutation-driven targeted therapies. Of these eight patients, four achieved partial response or stable disease for at least 4 months, and three were not assessable for response due to short exposure. There were no biopsy-related complications requiring additional treatment. CONCLUSION: Our findings suggest that US-guided liver tumor biopsy is a useful and safe method for obtaining high-quality samples for NGS-based clinical sequencing. In cases with metastatic liver tumors, US-guided biopsy should be considered to provide accurate and optimal sequencing results for patients.