The main goal of Crohn's disease (CD) treatment at present is to induce and maintain remission for as long as possible, and several approaches have been used as induction and maintenance therapies. There are no reports that have compared the effects on mid- and long-term prognosis among the induction and maintenance therapies, especially between infliximab, a chimeric antibody to tumor necrosis factor-alpha, and nutritional therapies. A total of 262 CD patients with induced remission were enrolled in the cohort study. Patients who failed to achieve remission, and patients who were lost to follow-up within 12 months were excluded. Induction therapies for CD included total elemental enteral nutrition, total parenteral nutrition, infliximab, prednisolone, and surgical resection. Maintenance therapies included home elemental diet, 5-aminosalicylates, immunomodulators, and scheduled infliximab therapy. We evaluated the possible predictive factors of relapse and surgical recurrence including the clinical backgrounds of the patients and medical therapies, using the Cox multivariate hazard analysis. The main factors that strongly affected the first relapse were scheduled infliximab therapy (hazard ratio (HR) = 0.24, p < 0.0001), surgical induction (HR = 0.19, p < 0.0001) and high frequency of previous relapse (HR = 2.56, p = 0.002). Penetrating (HR = 3.33, p = 0.009) and stricturing (HR = 6.60, p < 0.0001) disease behavior were main risk factors of surgical recurrence. Scheduled infliximab therapy is the most effective maintenance therapy in a real clinical setting with respect to the mid- and long-term prognosis.