Abstract Background Anti-TNF antibodies treatments are the only first-line reimbursed biologics for Crohn’s disease (CD) in several countries. Recently, Vedolizumab (VDZ) and Ustekinumab (UST) were added to the therapeutic armamentarium for CD refractory to a first anti-TNF antibody. However, studies comparing these two compounds remain unavailable. Our aim was to compare their efficacy in second-line treatment in CD after failure of one TNF antagonist. Methods All patients with CD refractory (primary or secondary non-responders) to first anti-TNF treatment and receiving UST or VDZ as a second biologic were included retrospectively in 10 French tertiary centres. The remission and clinical response were assessed at week 14. On the long-term, the cumulative probabilities of being in remission were estimated using the Kaplan–Meier method and the associated factors using a Cox proportional risk model. The drug survival to assess efficacy as well as side effects was assessed by actuarial analysis. Results 88 patients were included, 50 (57%) females (mean age: 41 ± 15 years), 61 (69%) being treated with UST and 27 (31%) with VDZ. The first anti-TNF was discontinued for primary or secondary non-response in 66 (75%) patients and for side effects in 22 (25%) patients, without any difference between the anti-TNF antibody previously used. Among the 55 patients with endoscopic data at baseline, 55 (98%) had ulceration, a CRP above 5mg/l for 33/71 (46%) patients and a faecal calprotectin > 250 µg/g for the 12 patients tested. At week 14, no difference was observed for clinical response and clinical remission according to the type of treatment: the rate of clinical response and remission were 74% (UST)/58% (VDZ) (p = 0.20) and 33% (UST)/26% (VDZ) (p = 0.56), respectively. The only factor associated with short-term remission was the lack of optimisation prior to anti-TNF discontinuation (p = 0.038) regardless of the type of second-line therapy (UST, p = 0.02; VDZ, p = 0.03). After a mean follow-up of 67 weeks, the cumulative probabilities of being in remission at 6 and 12 months were 16% and 34% for UST and 25% and 44% for VDZ, respectively (p = 0.24 for UST vs. VDZ). The drug survival was higher in the UST group as compared with the VDZ group (HR (UST vs. VDZ) = 2.36 [1.02–5.5], p = 0.04). Conclusion Our preliminary results suggest that VDZ and UST have similar efficacy in the short- and long-term response when used as a second-line biologic therapy in CD refractory to a first anti-TNF antibody. These results will be complemented for the congress by the inclusion of additional patients recruited into this registry.