Studies about long-term entecavir (ETV) therapy for partial virological response (PVR) are lacking. This study aimed to assess the clinical course of PVR patients receiving ETV therapy and analyze the efficacy of tenofovir (TDF). We retrospectively evaluated 130 patients who showed a PVR to ETV. Among these patients, 102 were nucleot(s)ide analogue (NUC)-naïve and 28 were lamivudine (LAM)-experienced. The cumulative rates of VR were 54.1%, 70.8%, and 83.7% for the NUC-naïve group and 37.0%, 42.8%, and 42.8% for the LAM-experienced group after 24, 36, and 48 months of ETV therapy, respectively (P = 0.008). Low HBV DNA level at 12 months (P < 0.001) and absence of a LAM treatment history (P = 0.031) were significant associated factors for VR. In VR prediction at 36 months of ETV therapy in NUC-naïve patients, HBV DNA level <95 IU/ml at 12 months showed a 92.9% sensitivity and a 78.3% specificity (AUROC, 0.909; P < 0.001). ETV resistance did not develop in NUC-naïve patients with HBV DNA levels <95 IU/ml at 12 months. The cumulative probability of VR in patients who switched to or additionally received TDF was 91.3% at 15 months. Prolonged ETV therapy induced a VR without the risk of ETV resistance in NUC-naïve patients with HBV DNA levels <95 IU/ml at 12 months. All patients with LAM-experienced or NUC-naïve with HBV DNA levels ≥95 IU/ml at 12 months should be switched to TDF rescue therapy. J. Med. Virol. 88:252–259, 2016. © 2015 Wiley Periodicals, Inc.