Background: There is uncertainty regarding the optimal duration of maintenance treatment with azathioprine (AZA) in Crohn's disease (CD) and ulcerative colitis (UC), with some studies suggesting durations up to 4–5 years and others supporting long-term indefinite use. More recent guidelines state that for patients in long term remission, cessation of treatment may be considered in the absence of objective signs of inflammation, but no data are available on how to monitor these patients after AZA withdrawal. Methods: A prospective observational study was performed on consecutive patients with CD or UC who stopped their maintenance treatment with AZA while being in steroid-free, deep remission, defined as normal clinical and endoscopic indexes, normal C-reactive protein (CRP) and normal fecal calprotectin. After AZA withdrawal, all patients received maintenance treatment with salicylates. Every 3 months, blood samples were collected for standard biochemical tests (including CRP, full blood count and protein electrophoresis), as well as stool samples for fecal calprotectin dosage. Bowel ultrasonography and ileocolonoscopy were performed every 6 and 12 months, respectively, as well as in case of clinical relapse. Multiple logistic regression and Cox proportional-hazards models were used to assess the predictors of disease relapse. Results: 41 patients (23 males, median age 45 years) who stopped AZA after median 7 years (range 5–14) were enrolled. After a mean follow-up of 24 months (range 5–48), 19 patients (46%; 7/19 CD, 12/22 UC) relapsed, within median 13 months (range 2–35). Predictors of clinical relapse at the time of AZA withdrawal were female gender (p=0.045) and age ≤45 years (p=0.035) only for UC. Fecal calprotectin was the only predictor of clinical relapse during the non invasive monitoring performed after AZA withdrawal, both in UC (p=0.019) and in CD (p=0.017). Moreover, fecal calprotectin was positive in all patients with UC and in 57% of CD patients at the time of relapse. Conclusions: Up to half of patients with IBD relapses after AZA withdrawal at different times. Fecal calprotectin can predict subsequent clinical relapse after AZA withdrawal. This can be useful in the non invasive monitoring of patients to detect preclinical relapse and to early correct our therapeutic strategy after AZA withdrawal. Prospective randomized studies are needed to validate such strategy.