Corticosteroids are still the first-line treatment for active ulcerative colitis more than 50 years after the publication of trials assessing their beneficial effect, with about a 50% remission rate in cases of severe disease. The mortality related to severe attacks of ulcerative colitis has decreased dramatically, to less than 1%, in experienced centers, due to the appropriate use of intensive therapeutic measures (intravenous steroids, fluids and electrolytes, artificial nutritional support, antibiotics, etc), along with timely decision-making about second-line medical therapy and early identification of patients requiring colectomy. One of the most difficult decisions in the management of severe ulcerative colitis is knowing for how long corticosteroids should be administered before deciding that a patient is a non-responder. Studies assessing the outcome of acute attacks after steroid initiation have demonstrated that, in steroid-sensitive patients, the response generally occurs early on, in the first days of treatment. Different indexes to predict treatment failure, when applied on the third day of treatment, have demonstrated a high positive predictive value for colectomy. In contrast to this resolute approach, which is the most widely accepted, other authors have suggested that in some patients a complete and prolonged response to steroids may take longer. Either way, physicians taking care of these patients need to recognize that severe ulcerative colitis may be life-threatening, and they need to be careful with excessively prolonged medical treatment and delayed surgery.