Sorafenib exerts modest antitumor activity in patients with advanced hepatocellular carcinoma (HCC), and radiological progressive disease (rPD) does not always correspond to so-called clinical progressive disease (cPD). We evaluated 101 patients who initiated sorafenib treatment for HCC and assessed post-progression survival (PPS) using the Cox proportional hazards model. PPS was calculated from the date of the first rPD until the date of death or the last follow-up. Using Cox model analysis of the 76 patients who experienced first rPD, we identified the Child-Pugh class, Eastern Cooperative Oncology Group performance status, the best antitumor response during treatment (using Response Evaluation Criteria in Solid Tumors (RECIST) Version 1.1) and α-fetoprotein levels as independent factors affecting PPS. When these factors were used to define scores ranging from zero to five with a cutoff value of two, PPS of patients who received best supportive care (BSC) after rPD was not statistically significantly different from that of patients who received post-rPD therapy with scores ≥2 (p = 0.220). In contrast, the PPS for the post-rPD therapy group was significantly longer compared with the BSC patients with scores <2 (p < 0.001). Patients who scored ≥2 at their first rPD were judged cPD and as candidates for BSC.