In central Taiwan, little is known about the different long‐term effects on renal function of the variant nucleoside analogs for chronic hepatitis B (CHB) patients. We aimed to compare the effects of long‐term use of different nucleoside analogs on renal function at a single medical center. We analyzed patients who fulfilled the criteria of the Bureau of National Health Insurance and received nucleoside analogs between October 1, 2003 and April 30, 2011. Patients were divided into three groups depending on the type of drug administered, which included lamivudine (LAM), entecavir (ETV), and telbivudine (LdT). We excluded CHB patients who received prophylaxis because of chemotherapy and those who underwent bone marrow or liver transplantation. The values of renal function at baseline before treatment and at one year after therapy were analyzed. A total of 307 patients were recruited, including 76 patients who were treated with LAM, 200 patients who were treated with ETV, and 31 patients who were treated with LdT. The initial estimated glomerular filtration rate (eGFR) (T0), as computed by the Modification of Diet in Renal Disease (MDRD) formula and expressed as mL min−1 1.73 m−2, was 86.56 ± 25.30 for LAM, 106.07 ± 36.91 for ETV, and 117.23 ± 42.70 for LdT. The mean change in eGFR (∆eGFR: T12‐T0) at one year after treatment (T12) from baseline (T0) was −3.55 ± 19.13 for LAM, −3.53 ± 22.85 for ETV, and 9.54 ± 29.93 for LdT. For all patients, the ∆eGFR was significantly different between the LdT group and the other two groups (P = 0.006), especially in the subgroups who had eGFR lower than 90 mL min−1 1.73 m−2 (P = 0.001). The type of nucleoside analog was independently associated with ∆eGFR (OR, 4.41; P < 0.001). Renal function improved in CHB patients who received LdT for 1 year compared with those who received LAM or ETV, especially in patients who had eGFR of less than 90 mL min−11.73 m−2.