Background: The investigation of glycated hemoglobin (HbA1c) as a diagnostic tool for abnormal glycometabolism is lack in acute ischemic stroke patients in China and worldwide. This paper was aimed to determine whether HbA1c, fasting plasma glucose (FPG), or HbA1c combined with FPG, could be used to screen for diabetes mellitus (DM) or prediabetes in acute ischemic stroke patients without previous DM. Methods: Acute ischemic stroke patients without previous DM (n = 1,316) were selected from the Abnormal gluCose Regulation in Patients with Acute StrOke acrosS China Study (ACROSS-China). Oral glucose tolerance test (OGTT), HbA1c, FPG, and HbA1c combined with FPG were used as the screening methods to categorize the glycometabolic status. OGTT was taken as the golden method. Venn diagrams and the overlap index were used to determine the associations among the three methods of identifying abnormal glycometabolism. The area under the receiver operating characteristic curve (AUROC) and Youden index were used to assess and compare the accuracy in detecting abnormal glycometabolism. Youden analyses were performed to determine the ideal cutoff values of HbA1c in diagnosing abnormal glycometabolism. Results: In acute ischemic stroke patients without previous DM, the overlaps of HbA1c versus OGTT, HbA1c versus FPG, and all the three methods independently, were low for detecting abnormal glycometabolism (all <50%). HbA1c can significantly detect more cases of prediabetes than OGTT (P < 0.001). The combination of HbA1c and FPG significantly raised the sensitivity to over 60.0%, specificity to over 80.0%, and the diagnostic accuracy (Youden index from under 40.0% to 42.4%) for DM. HbA1c of 5.7%-6.4% had a low to moderate concordance with OGTT for identifying prediabetes (AUROC = 0.557, P = 0.001). HbA1c values of 6.3% and 5.9% were found to be the ideal cutoff values for detecting DM and abnormal glycometabolism in our data, respectively. Conclusions: The combination of HbA1c and FPG increased the diagnostic rate of DM when compared with OGTT, and increased the diagnostic accuracy for DM compared with HbA1c or FPG alone. Our results advocate the use of HbA1c as screening tool for the diagnosis of pre-diabetes.