Objective: To explore the status of adolescents’ health risk behaviors, relationship between cognitive ability and health risk behaviors, and provide an empirical reference for reducing the occurrence and early prevention of health risk behaviors among adolescents. Methods: Samples were extracted by hierarchical, multi-stage, probability and scale proportional (PPS) methods. There were 112 schools randomly selected from 28 county-level regions(county, district, city), and 19 487 students were investigated by baseline survey, a follow-up study was conducted among 8 636 students during 2014-2015. Descriptive statistical methods of using 2 test and t test, and logistic multivariate regression analysis were used to analyze the relationship between cognitive abilities and healthy risk behaviors. Results: The reported rate of a healthy risk behavior was 20.77%, and the incidence of more than one type of health risk behavior was 16.03%. The student’s cognitive ability standardized score was(0.33±0.82). The average score of cognitive ability of adolescents with healthy risk behavior was 0.14, which was lower than the average score of cognitive ability of adolescents without healthy risk behaviors(0.43). Boys, having rural household registeration, non-residential schools, and parents with low education levels adolescents were more likely to be involved in health behavior risks. There was a significant correlation between adolescent cognitive ability and healthy risk behavior(OR=0.67, P<0.01). Conclusion The current status of adolescent health risk behaviors is worrying. The cognitive ability has a significant impact on health risk behaviors. It is recommended to use the cognitive ability of adolescents as a starting point to prevent and reduce the occurrence of adolescent health risk behaviors.