Background/Aims: In patients with active ulcerative colitis (UC), pharmacologics, although initially effective in most patients, are associated with refractoriness, loss of response or unfavourable side effects as additional morbidity factors. Depletion of myeloid lineage leucocytes like the CD14(+)CD16(+) monocyte phenotype, which is a major source of tumour necrosis factor-α, by granulocyte/monocyte apheresis (GMA) if effective, is also known to be free from side effects. Methods: In clinical practice setting, 77 consecutive patients with moderate to severe UC, who failed to respond to first-line medications received GMA with the Adacolumn as remission induction therapy. Patients who achieved remission were followed for 3 years. Results: Among the 77 patients, 46.8% were corticosteroids-naïve, 26% corticosteroid-dependent and 27.3% corticosteroid-refractory. The overall clinical remission rate was 79.2%, and the overall mucosal healing (MH) rate according to the Mayo endoscopic subscore ≤1 was 58.5%. MH rates in corticosteroid-naïve, corticosteroid-dependent and corticosteroid-refractory subgroups were 70.8, 56.3, and 38.5%, respectively. The 3-year sustained clinical remission rates in corticosteroid-naïve, corticosteroid-dependent and corticosteroid-refractory subgroups were 83.3, 68.8, and 23.1%, respectively. Conclusion: Corticosteroid-naïve patients appeared to benefit the most from the Adacolumn GMA, and attain a favourable long-term clinical course. Accordingly, GMA should be a first-line therapy in this clinical setting.