Summary Background Controversy exists regarding glucocorticoids therapy and the risk of peptic ulcer bleeding (PUB). Aim The present study was undertaken to determine whether short‐term use of glucocorticoids is associated with the occurrence of peptic ulcer bleeding. Methods The records of adult patients hospitalised for newly diagnosed peptic ulcer bleeding from 2000 to 2012 were retrieved from the Taiwan National Health Insurance Research Database, a nationwide population‐based registry system. The association between systemic glucocorticoids usage and peptic ulcer bleeding was determined with a conditional logistic regression model comparing cases and controls during time windows of 7, 14 and 28 days using a case‐crossover design. Results Of the 8894 enrolled patients, the adjusted self‐matched odds ratios for peptic ulcer bleeding after exposure to the glucocorticoids were 1.37 (95% CI: 1.12–1.68, P = 0.003) for the 7‐day window, 1.66 (95% CI: 1.38–2.00, P < 0.001) for the 14‐day window and 1.84 (95% CI: 1.57–2.16, P < 0.001) for the 28‐day window. Moderate to high, but not low dose glucocorticoids (methylprednisolone <4 mg/day or its equivalence) were associated with an increased risk of peptic ulcer bleeding. Concomitant use of a nonselective nonsteroidal anti‐inflammatory drug (NSAID) or aspirin further elevated the risk. However, it does not eliminate the effect of underlying diseases flare‐up that may have placed the patients at risk for peptic ulcer bleeding in this kind of study design. Conclusions Short‐term (7–28 days) exposure to glucocorticoids is significantly associated with peptic ulcer bleeding; this risk seems dose‐dependent and is higher when nonselective NSAIDs or aspirin are used concurrently.