Background Treatment failure occurs in 20% of autoimmune hepatitis patients on prednisolone and azathioprine (AZA). There is no established second line treatment. Aim To assess the efficacy of mycophenolate mofetil as second line treatment after AZA-intolerance or AZA-nonresponse in autoimmune hepatitis and overlap syndromes. Methods Consecutive patients from the Dutch Autoimmune Hepatitis Group cohort, consisting of 661 patients, with autoimmune hepatitis or overlap syndromes, AZA-intolerance or AZA-nonresponse and past or present use of mycophenolate mofetil were included. Primary endpoint of mycophenolate mofetil treatment was biochemical remission. Secondary endpoints were biochemical response (without remission), treatment failure and prevention of disease progression. Results Forty-five patients treated with mycophenolate mofetil were included. In autoimmune hepatitis remission or response was achieved in 13% and 27% in the AZA-nonresponse group compared to 67% and 0% in the AZA-intolerance group (P = 0.008). In overlap-syndromes remission or response was reached in 57% and 14% in the AZA-nonresponse group and 63% and 25% of the AZA-intolerance group (N.S.); 33% had side effects and 13% discontinued mycophenolate mofetil. Overall 38% had treatment failure; this was 60% in the autoimmune hepatitis AZA-nonresponse group. Decompensated liver cirrhosis, liver transplantations and death were only seen in the autoimmune hepatitis AZA-nonresponse group (P < 0.001). Conclusions Mycophenolate mofetil induced response or remission in a majority of patients with autoimmune hepatitis and azathioprine-intolerance and with overlap syndromes, irrespective of intolerance or nonresponse for azathioprine. In autoimmune hepatitis with azathioprine nonresponse mycophenolate mofetil is less often effective.