Summary Background Aetiological therapy improves liver function and may enable hepatic recompensation in decompensated cirrhosis. Aims We explored the potential for recompensation in patients with decompensated primary biliary cholangitis (PBC) – considering a biochemical response to ursodeoxycholic acid (UDCA) according to Paris‐II criteria as a surrogate for successful aetiological treatment. Methods Patients with PBC were retrospectively included at the time of first decompensation. Recompensation was defined as (i) resolution of ascites and hepatic encephalopathy (HE) despite discontinuation of diuretic/HE therapy, (ii) absence of variceal bleeding and (iii) sustained liver function improvement. Results In total, 42 patients with PBC with decompensated cirrhosis (age: 63.5 [IQR: 51.9–69.2] years; 88.1% female; MELD‐Na: 13.5 [IQR: 11.0–15.0]) were included and followed for 41.9 (IQR: 11.0–70.9) months after decompensation. Seven patients (16.7%) achieved recompensation. Lower MELD‐Na (subdistribution hazard ratio [SHR]: 0.90; p = 0.047), bilirubin (SHR per mg/dL: 0.44; p = 0.005) and alkaline phosphatase (SHR per 10 U/L: 0.67; p = 0.001) at decompensation, as well as variceal bleeding as decompensating event (SHR: 4.37; p = 0.069), were linked to a higher probability of recompensation. Overall, 33 patients were treated with UDCA for ≥1 year and 12 (36%) achieved Paris‐II response criteria. Recompensation occurred in 5/12 (41.7%) and in 2/21 (9.5%) patients with vs. without UDCA response at 1 year, respectively. Recompensation was linked to a numerically improved transplant‐free survival (HR: 0.46; p = 0.335). Nonetheless, 4/7 recompensated patients presented with liver‐related complications after developing hepatic malignancy and/or portal vein thrombosis and 2 eventually died. Conclusions Patients with PBC and decompensated cirrhosis may achieve hepatic recompensation under UDCA therapy. However, since liver‐related complications still occur after recompensation, patients should remain under close follow‐up.