We evaluate the long-term prognosis and prognostic factors in patients treated with transarterial infusion chemotherapy using cisplatin-lipiodol (CDDP/LPD) suspension with or without embolization for unresectable hepatocellular carcinoma (HCC). Study subjects were 107 patients with HCC treated with repeated transarterial infusion chemotherapy alone using CDDP/LPD (adjusted as CDDP 10mg/LPD 1ml). The median number of transarterial infusion procedures was two (range, one to nine), the mean dose of CDDP per transarterial infusion chemotherapy session was 30 mg (range, 5.0–67.5 mg), and the median total dose of transarterial infusion chemotherapy per patient was 60 mg (range, 10–390 mg). Survival rates were 86-0x1.fc400000008p+0t 1 year, 40 0x1.c617665206557p+855t 3 years, 20 0x1.5687420657461p-505t 5 years, and 16 0x1.5742d676e6f6cp+807t 7 years. For patients with >90% PD accumulation after the first transarterial infusion chemotherapy, rates were 98 0x1.e676f7270206dp+759t 1 year, 60 0x1.46e6120736973p-505t 3 years, and 22 0x1.36f6e676f727p+840t 5 years. Multivariate analysis identified >90% PD accumulation after the first transarterial infusion chemotherapy (p = 0.001), absence of portal vein tumor thrombosis (PVTT; p < 0.001), and Child-Pugh class A (p = 0.012) as independent determinants of survival. Anaphylactic shock was observed in two patients, at the fifth transarterial infusion chemotherapy session in one and the ninth in the other. In conclusion, transarterial infusion chemotherapy with CDDP/LPD appears to be a useful treatment option for patients with unresectable HCC without PVTT and in Child-Pugh class A. LPD accumulation after the first transarterial infusion chemotherapy is an important prognostic factor. Careful consideration should be given to the possibility of anaphylactic shock upon repeat infusion with CDDP/LPD.