Although systemic steroids remain as the gold standard for the treatment of acute moderate to severe active ulcerative colitis (UC), 15–57% of patients do not achieve clinical remission. We sought to identify clinical, biological, or radiologic predictive factors of response to steroid treatment in acute UC attacks. The medical records of 39 consecutive patients admitted for an acute attack of UC and treated with systemic steroids, were reviewed. Epidemiologic, demographic, and clinical data at baseline and clinical data 3 days after starting steroid treatment were registered. Treatment failure was defined as the need of IV cyclosporine or colectomy before hospital discharge. Twenty-four patients (62%) responded to systemic steroids. Thirteen out of the 15 nonresponders, were treated with IV cyclosporine, avoiding colectomy in 7 cases (54%). More than six bowel movements per day at the third day of treatment, blood in stools in the third day of therapy, extensive UC, and the presence of malnutrition were associated with steroid treatment failure, but only blood in stools (P = .04), and more than six movements per day (P = .012) after 3 days of treatment, were found to be independent predictive factors of steroid refractoriness. In conclusion, clinical evaluation 3 days after starting systemic steroids seems to be the best tool to assess short-term prognosis.