BACKGROUND: Adherence to medications is important to maintain disease under control and to prevent complications in pregnant patients with ulcerative colitis (UC). To evaluate the incidence of non-adherence during pregnancy and its effect on relapse and pregnancy outcomes, we conducted a multicenter prospective study using a patient self-reporting system without physician interference. METHODS: Sixty-eight pregnant UC women were recruited from 17 institutions between 2013 and 2019. During the course of pregnancy, questionnaires were collected separately from patients and physicians, to investigate the true adherence to medications, disease activity, and birth outcomes. Multivariable logistic regression analysis was performed to identify the risk factors for the relapse or adverse pregnancy outcomes. RESULTS: Of 68 pregnancy, 15 adverse pregnancy outcomes occurred in 13 patients. The rate of self-reported non-adherence was the greatest to mesalamines in the first trimester, which was significantly higher than physicians' estimate (p = 0.0116), and discontinuation was observed in 42.1% of non-adherent group. Logistic regression analysis revealed non-adherence as an independent risk factor for relapse [odds ratio (OR) 7.659, 95% CI 1.928-30.427, p = 0.038], and possibly for adverse pregnancy outcome (OR 8.378, 95% CI 1.350-51.994, p = 0.023). Among the subgroup of patients treated with oral mesalamine alone, the non-adherence was confirmed to be an independent risk factor for relapse (p = 0.002). CONCLUSION: Non-adherence to mesalamine was underestimated by physicians in pregnant UC patients and contributed to disease relapse and possibly on pregnancy outcomes. Preconceptional education regarding safety of medications and risk of self-discontinuation is warranted.