Primary liver cancer (hepatocellular carcinoma; HCC) in patients not eligible for surgery or transplant is currently treated by locoregional therapeutic approaches, including trans-arterial chemoembolization and radiofrequency ablation. Sorafenib (Nexavar; Bayer/Onyx) is currently the only approved systemic therapy for patients having failed locoregional interventions. Oncolytic viruses are designed to selectively replicate within, and subsequently lyse, cancer cells by a unique mechanisms-of-action that is not cross-resistant with approved therapies (Kirn et al., Nat Med 7:781–787, 2001; Parato et al., Nat Rev Cancer 5:965–976, 2005; Chiocca, Nat Rev Cancer 2:938–950, 2002; Heise and Kern, J Clin Invest 105:847–851, 2000). Given that these therapeutics are self-amplifying in tumors, the impact of dose on patient outcome is unclear. Pexa-Vec (JX-594) is an oncolytic and immunotherapeutic vaccinia virus which was shown to be well tolerated by intratumoral injection and intravenous infusions in Phase 1 trials (Park et al., Lancet Oncol 9:533–542, 2008; Breitbach et al., Nature 477:99–102, 2011). We present the design of a randomized dose-finding trial of Pexa-Vec in patients with advanced HCC in which Pexa-Vec was delivered by intratumoral injection three times every 2 weeks at one of two dose levels (1 × 10⁸ plaque forming units (pfu) versus 1 × 10⁹ pfu).