Background Although cardiac troponin has been well established as diagnostic and prognostic makers for acute coronary heart disease, the prognostic value of elevated cardiac troponin in patients with intracerebral hemorrhage (ICH) was inconsistent and not systematically evaluated. Hypothesis We proposed the hypothesis that the practical utility of cardiac troponin levels for prediction of mortality and poor outcome in ICH patients. Methods A total of 1004 patients with ICH were retrospectively reviewed and qualified for further analysis from June 2012 to December 2015. The patients were divided into different groups based on measurements of cardiac troponin I (cTnI) at the time of admission and the following day. Multivariate Cox proportional hazards analysis were performed to determine the independent prognostic value of the cTnI for patients in-hospital mortality and poor outcomes, the receiver operator characteristic (ROC) analysis was performed to assess the predictive value of cTnI, ICH score, and combination of them. Results Serum cTnI level was an independent predictor in-hospital mortality (positive vs negative, HR (hazard ratios) = 3.44, 95% CI (confidence interval) 1.66-7.13, P < .001) and poor outcomes in patients with ICH (positive vs negative, HR = 6.69, 95% CI 4.25-10.52, P < .001). Addition of cTnI to ICH score significantly improved the prognostic discrimination for both in-hospital mortality and poor outcomes. Conclusion Serum cTnI levels may be valuable as predictor for in hospital mortality and poor outcomes and may be useful in the risk stratification of ICH during hospitalization.