ObjectiveTo compare the effect of different indicators on stress‐induced hyperglycemia for predicting in‐hospital outcomes of acute intracerebral hemorrhage.MethodsUsing data from the Chinese Stroke Center Alliance database, which is a national, multicenter, prospective, and consecutive program. Stress‐induced hyperglycemia was described as glycemic gap (GG, defined as fasting blood glucose [FBG] minus estimated average blood glucose) and stress hyperglycemia ratio (SHR, defined as FBG‐to‐estimated average blood glucose ratio [SHR 1] or FBG‐to‐HbA1c ratio [SHR 2]). The primary outcome was in‐hospital mortality, and the second outcome was hematoma expansion.ResultsA total of 71,333 patients with acute intracerebral hemorrhage were included. In multivariate analyses, the highest levels of GG (OR 1.68, 95% CI 1.12–2.51), SHR 1 (OR 1.73, 95% CI 1.15–2.60), and SHR 2 (OR 2.07, 95% CI 1.33–3.23) were associated with in‐hospital death (all the p trends <0.01). Only the highest level of SHR 2 (OR 1.24 [1.02–1.51], p trend >0.05) was related to hematoma expansion. No association between GG or SHR 1 and hematoma expansion was observed. The areas under the ROC curve of GG, SHR 1, and SHR 2 for in‐hospital mortality were 0.8808 (95% CI 0.8603–0.9014), 0.8796 (95% CI 0.8589–0.9002), and 0.8806 (95% CI 0.8600–0.9012). The areas under the ROC curve of SHR 2 for hematoma expansion were 0.7133 (95% CI 0.6964–0.7302).InterpretationSHR (FBG‐to‐HbA1c ratio) was associated with both in‐hospital death and hematoma expansion in intracerebral hemorrhage, and might serve as an accessory indicator for the in‐hospital prognosis of intracerebral hemorrhage.