Although anti-tumor necrosis factor-α monoclonal antibody biological preparations (BP) agents are widely used as an established treatment tool for refractory ulcerative colitis (UC), whether leukocytapheresis/granulocytapheresis (L/G-CAP) has similar beneficial impact on the disease activity remains undetermined. Furthermore, the costs defrayed for the treatment with these 2 modalities have not been compared. We retrospectively evaluated whether L/G-CAP offered sustained beneficial effects over 2-year period. The patients who had moderately to severely active UC (Rachmilewitz clinical activity index (CAI) ≧ 5) and were treated with a series (10 sessions) of L/G-CAP (n = 19) or BP (n = 7) as an add-on therapy to conventional medications were followed. Furthermore, the cost-effectiveness pertaining to the treatment with L/G-CAP and BP was assessed over 12 months. At baseline, L/G-CAP and BP groups manifested similar disease activity (CAI, L/G-CAP; 7.0 [6.0-10.0], BP; 10.0 [6.0-10.0], P = .207). The L/G-CAP and BP treatment suppressed the activity, with CAI 1 or less attained on day 180. When the L/G-CAP group was dichotomized into L/G-CAP-high and L/G-CAP-low group based on CAI values (≥3 or < 3) on day 365, CAI was gradually elevated in L/G-CAP-high group but remained suppressed in L/G-CAP-low group without additional apheresis for 2 years. Anemia was corrected more rapidly and hemoglobin levels were higher in BP group. The cost of the treatment with L/G-CAP over 12 months was curtailed to 76% of that with BP (1.79 [1.73-1.92] vs 2.35 [2.29-3.19] million yen, P = .028). L/G-CAP is as effective as BP in a substantial number of patients over 2 years. The cost for the treatment of UC favors L/G-CAP although the correction of anemia may prefer BP. Thus, L/G-CAP can effectively manage the disease activity with no additional implementation for 2 years although further therapeutic modalities might be required in a certain population with high CAI observed on day 365.