Many patients with moderately to severely active Crohn's disease do not respond to available therapies or lose response over time. The GALAXI-1 study previously found that three intravenous guselkumab dosages showed superior clinical and endoscopic outcomes over placebo at week 12 in patients with moderately to severely active Crohn's disease. We report the safety and efficacy of subcutaneous guselkumab maintenance regimens to week 48 in the GALAXI-1 study. We did a phase 2, randomised, multicentre, double-blind trial. Adult patients with moderately to severely active Crohn's disease were randomly allocated with a computer-generated randomisation schedule to receive one of five treatment groups, with regimens consisting of an intravenous induction phase transitioning to a subcutaneous maintenance phase starting at week 12 in a treat-through design: (1) guselkumab 200→100 mg group (200 mg intravenous at weeks 0, 4, and 8, then 100 mg subcutaneous every 8 weeks; (2) guselkumab 600→200 mg group (600 mg intravenous at weeks 0, 4, and 8, then 200 mg subcutaneous every 4 weeks); (3) guselkumab 1200→200 mg group (1200 mg intravenous at weeks 0, 4, and 8, then 200 mg subcutaneous every 4 weeks); (4) ustekinumab group (approximately 6 mg/kg intravenous at week 0, then 90 mg subcutaneous every 8 weeks); or (5) placebo group (placebo induction followed by either placebo maintenance [for those with CDAI clinical response at week 12] or crossover to ustekinumab [for those without CDAI clinical response at week 12]). Endpoints assessed at week 48 included CDAI remission (CDAI score <150), endoscopic response (≥50% improvement from baseline in SES-CD or SES-CD score ≤2), and endoscopic remission (SES-CD score ≤2) in the primary efficacy analysis population of all randomised patients who received at least one dose of study drug, excluding those discontinued during a temporary study pause. Safety analyses included all randomised patients who received at least one study drug dose. This trial is registered at Clinical Trials.gov (NCT03466411) and is active but not recruiting. Among 700 patients screened, 309 (112 biologic-naive; 197 biologic-experienced) were included in the primary efficacy analysis population: 61 in the guselkumab 200→100 mg group, 63 in the guselkumab 600→200 mg group, 61 in the guselkumab 1200→200 mg group, 63 in the ustekinumab group, and 61 in the placebo group. 126 (41%) women and 183 (59%) men were included, with median age 36·0 years (IQR 28·0-49·0). At week 48, the numbers of patients with CDAI clinical remission were 39 (64%) in the guselkumab 200→100 mg group, 46 (73%) in the guselkumab 600→200 mg group, 35 (57%) in the guselkumab 1200→200 mg group, and 37 (59%) in the ustekinumab group. The corresponding numbers of patients with endoscopic response were 27 (44%), 29 (46%), 27 (44%), and 19 (30%), respectively, and endoscopic remission was seen in 11 (18%), 11 (17%), 20 (33%), and four (6%) patients, respectively. In the placebo group, 15 patients were in CDAI clinical response at week 12 and continued placebo; of these, nine (60%) were in clinical remission at week 48. 44 patients in the placebo group were not in CDAI clinical response at week 12 and crossed over to ustekinumab; of these, 26 (59%) were in clinical remission at week 48. Up to week 48, adverse events frequencies in the safety population (n=360) were 46 (66%) of 70 patients (464·9 events per 100 patient-years of follow-up) in the placebo group, 163 (74%) of 220 patients (353·1 per 100 patient-years) in the three guselkumab groups combined, and 60 (85%) of 71 patients (350·7 per 100 patient-years) in the ustekinumab group. Among patients treated with guselkumab or ustekinumab, the most frequently reported infections up to week 48 were nasopharyngitis (25 [11%] of 220 guselkumab recipients, 12 [11%] of 114 ustekinumab recipients) and upper respiratory infections (13 [6%] guselkumab recipients, eight [7%] ustekinumab recipients). After week 12, one patient who responded to placebo induction and two guselkumab-treated patients had serious infections. No active tuberculosis, opportunistic infections, or deaths occurred. Patients receiving guselkumab intravenous induction and subcutaneous maintenance treatment achieved high rates of clinical and endoscopic efficacy up to week 48. No new safety concerns were identified. Janssen Research & Development.