Summary Background  Ulcerative colitis (UC) can be maintained in remission with 5‐aminosalicylic acid (5‐ASA) medications, but frequent non‐adherence by patients who are feeling well has been associated with more frequent flares of colitis. Aim  To perform a systematic review of the published literature and unpublished randomized clinical trials (RCTs) regarding the impact of non‐adherence with 5‐ASA medications on the incidence of UC flares and costs of care. Methods  A search of MEDLINE, EMBASE and the Cochrane databases was performed. Prospective studies of UC maintenance with 5‐ASAs in adults were selected if they included data on adherence and disease flares. Studies using insurance claims data to estimate the impact of non‐adherence on cost of care were included. Data from unpublished RCTs were obtained from the FDA with a request under the Freedom of Information Act. Results  The relative risk for flare in non‐adherent vs. adherent patients ranged from 3.65 to infinity. Data were obtained from six unpublished 5‐ASA RCTs, but none measured the impact of adherence on disease activity. The comorbidity‐adjusted annual costs of care in adherent patients were 12.5% less than in non‐adherent patients, despite increased medication expenditures. Conclusions  A substantial proportion of UC flares and medical costs of UC are attributable to 5‐ASA non‐adherence. As non‐adherence to 5‐ASA medications is common, cost‐effective strategies to improve adherence are needed. The impact of adherence on disease activity should be measured in RCTs of all inflammatory bowel disease treatments.