Corticosteroids are sometimes used in inoperable malignant bowel obstruction (MBO) despite limited evidence. The main randomised evidence comes from an old Cochrane review totalling 89 patients demonstrating a non-significant trend towards obstruction resolution and no survival benefit. Subsequent literature consists of retrospective and uncontrolled studies failing to control for spontaneous resolution, disease heterogeneity and concomitant therapies. Dose, duration and outcomes are inconsistently reported. Corticosteroid toxicities are potentially harmful. Overall, there are no high-quality data to support corticosteroids as a standard therapy in MBO. If a clinician wishes to use in selected patients, this should be as a time-limited therapeutic trial framed within the context of uncertainty.