Background Discontinuation of antitumour necrosis factor (TNF)alpha therapy with perianal fistulising Crohn's disease remains controversial due to the risk of severe relapse without any clear evidence. Aim The aim of this study was to assess the rate and type of perianal and luminal relapses following anti-TNF alpha discontinuation. Methods All patients treated with anti-TNF alpha for perianal fistulising Crohn's disease with subsequent discontinuation of therapy were retrospectively reviewed from a prospective database (1998-2016). Cumulative probabilities of relapse-free survival were estimated by actuarial analysis. Results After a median follow-up of 62 months, 24 of the 45 patients experienced perianal relapse. A new surgical drainage was needed in 19 (79%) patients. The cumulative probabilities of perianal relapse at 1 and 5 years were 24% and 55%, respectively. Ileal localization (L1) at diagnosis, persistence of an external fistula opening, second line anti-TNF alpha use, or prior dose optimization was associated with perianal relapse, whereas continuation of immunosuppressive agents decreased this risk (HR = 0.3). Luminal relapse occurred in 42% of patients at 5 years. The cumulative probability of global relapse at 5 years was 67%. Retreatment with anti-TNF alpha allowed further remission in 23 of 24 (96%) patients. Conclusion Half of patients with perianal fistulising Crohn's disease relapse within 5 years after anti-TNFa discontinuation. Immunosuppressant continuation may decrease this risk. The high risk of relapse (perianal and luminal) may suggest a benefit in pursuing biologics over a longer period in patients with perianal fistulas.