Aims Radiofrequency catheter ablation (RFCA) is now an established therapeutic option for patients with atrial fibrillation (AF), but the long-term recurrence rate of AF is still high. Sacubitril/valsartan (Sac/Val) is superior to valsartan in attenuating ventricular remodelling and improving clinical outcomes in heart failure patients, but whether this additional benefit exists in reversing atrial remodelling and reducing AF recurrence of RFCA-treated AF patients remains uncovered. Methods and results Patients that had undergone RFCA were enrolled and randomly assigned 1:1 to valsartan (160 mg/day) or Sac/Val (200 mg/day) treatment group, in addition to other standard treatment of AF. Patients were followed up for 24 weeks. Echocardiography and ambulatory Holter monitoring for 24 h was performed at 24 weeks after RFCA. The primary end point was the change of atrial diameter from baseline to 24 weeks after RFCA. Second end points included the recurrence rate of AF, all-cause hospitalization and all-cause death. A total of 64 AF patients were enrolled, 32 of which received Sac/Val and 32 received valsartan treatment. There was no difference in the age (64.8 +/- 9.8 vs. 63.7 +/- 9.0, P = 0.634), gender (per cent of male: 59.4% vs. 50.0%, P = 0.616), heart rate (84.7 +/- 4.1 b.p.m. vs. 80.9 +/- 2.6 b.p.m., P = 0.428), systolic (127.5 +/- 15.4 mmHg vs. 130.0 +/- 17.8 mmHg, P = 0.549) or diastolic (81.7 +/- 9.8 mmHg vs. 79.9 +/- 12.6, P = 0.537) blood pressure upon admission between valsartan and Sac/Val treatment groups. The percentage of persistent AF was also comparable (43.8% vs. 53.1%, P = 0.617) in both treatment groups. Patients receiving Sac/Val treatment displayed significant decrease in the left atrial diameter (4.3 +/- 0.5 cm to 3.8 +/- 0.5 cm, P < 0.001), volume index (48.0 +/- 6.4 mL/m(2) to 41.7 +/- 7.0 mL/m(2), P < 0.001), and right atrial diameter (4.4 +/- 0.8 cm to 3.9 +/- 0.7 cm, P = 0.017) from baseline to 24 weeks after RFCA. This effect was not observed in valsartan treatment group. There was a numerical decrease in AF recurrence rate in the Sac/Val group compared with valsartan group (9.4% vs. 15.6%), although this difference did not reach a statistical significance (P = 0.708). No difference in all-cause hospitalization rate (6.3% in each group) or all-cause death rate (0% in each group) was observed. Conclusions Our data indicate that Sac/Val is superior to valsartan in attenuating atrial structural remodelling in catheter ablation-treated AF patients.