The aim of this study was to investigate the effects of double‐filtration plasmapheresis (DFPP), immunoadsorption (IA) and intravenous immunoglobulin (IVIg) in the treatment of late‐onset myasthenia gravis (MG). A total of 40 late‐onset MG patients were randomly divided into three groups: 15 patients were treated with DFPP; 10 patients were treated with IA; and 15 patients received IVIg. The titers of titin antibodies (Titin‐ab), acetylcholine receptor antibodies (AChR‐ab), presynaptic membrane antibody (Prsm‐ab) were detected before and after the treatment, and the quantitative MG score (QMG score) was assessed by blinded examiners before and immediately after the entire course of treatment. The clinical efficacy, duration of respiratory support, hospital stay, and the correlation between the three antibodies and the QMG score were also analyzed. Compared to pre‐treatment, the values of Titin‐ab, AChR‐ab, and PrsmR‐ab were all dramatically decreased (P < 0.05); meanwhile the value of Titin‐ab in the DFPP and IA groups decreased much more than in the IVIg group (P < 0.01); however, no statistical difference was found between the DFPP and IA groups (P > 0.05). Although the QMG score significantly improved in all three groups, it decreased much more in both the DFPP and IA groups than that in the IVIg group (P < 0.01). Symptoms were also effectively ameliorated by all treatments, but the clinical efficacy of the DFPP and IA groups was higher than the IVIg group (P < 0.05), as was the remission time (P < 0.01), the duration of hospital stay (P < 0.05), and the number of respiratory supports required (P < 0.05). Using Pearson's correlation, the decrease of Titin‐ab showed a longitudinal correlation with the decrease of QMG score (r = 0.6107, P < 0.01). Both DFPP and IA showed better short‐term clinical effectiveness than immunoglobulin transfusion, rapidly and effectively clearing the pathogenic antibodies in late‐onset MG patients, especially for Titin‐ab.