Background: The motivation of the present study is to explore potential relationships between small molecule metabolites and animal behavior phenotype by integrating data of rat gait analysis with blood NMR metabonomics data.Methods and Results: (1)Central nervous system diseases characterized in cerebral blood flow reduction are adopted in this research.Two kinds of data are included: blood metabonomics data collected from rats suffering in middle cerebral artery occlusion (MCAO) and phenotypic data concerning animals behavior in gait analysis experiment.(2)Indicators like stance time, stride time, homolateral coupling are significantly increased in cerebral ischemia model group vs.control, and indicators like speed, percentage of swing are decreased.Cerebral ischemia model group and control group can also be clearly distinguished by 1 H-NMR blood metabonomics signatures.(3)Similarities of multidimensional phenotypic indicators are interrogated by hierarchical clustering analysis.All 90 behavioral indexes were clustered into two groups, and a group of representative index was identified.(4)Significant correlations were identified in multiple metabolites-behavioral index pairs.For example, some behavioral indexes like speed or stance time can be predicted by glucose concentration.Conclusion: The integrated analysis revealed that the concentrations of some small molecule metabolites in the blood are significantly associated with animal behavior phenotypes.Considering the regular animal behavioral measurements are time and resources consuming, the metabonomics features might be used as promising surrogate approach for evaluating the animal behavioral phenotype in neuropsychiatric drug evaluation .