Objectives: Less ischemic volume and better functional outcome are related with good cerebral collaterals in ischemic stroke. But the method of assessing collateral circulation is still undetermined. Methods: Collateral status was assessed using a computed tomography angiography (CTA) source image (CTA-SI) based and digital subtraction angiography (DSA) based assessment system. CTP maps were quantitatively analyzed. CBF, CBV, MTT, and TTP were calculated within the hypoperfusion and in the equivalent region of unaffected hemisphere. ASTIN-ASN system based on DSA and leptomeningeal collaterals (LMC) system based on CTA were adopted separately, and the intra-rater and inter-rater were determined by Kappa testing. Receiver Operating Characteristic (ROC) curves were used to analyze the relationship between CTP quantitative parameters and DSA. The related factors of collateral flow were also explored in this study. Results: Of 52 patients, 12(23.1%) were women, and the average age was 59.4y. The kappa-coefficients were 0.794 and 0.890 in the CTA and DSA collateral grade systems, respectively. The kappa-coefficients were 0.679 comparing with the CTA and DSA collateral grade systems. By means of ROC curves, a relative MTT (rMTT) was obtained as relative variables of collaterals. The cut-off value is 1.37 emerged as the only predictor of good collaterals. The multi-mode CT collaterals scores showed excellent consistency comparing with DSA (kappa-coefficients 0.770). Based on univariate and multivariate analysis, the rate of stenosis is the only predictor in the development of collateral circulation. Discussion: Our data showed that multi-mode CT score has a high value in predicting leptomeningeal collateral flow in ischemic stroke.