Risks brought by use and release of chemical substances to human health and the environment have caught more and more attention of government and research groups. It is impossible and unreasonable to conduct full assessment for all of the chemicals. For this reason, a number of screening prioritization exercises has been conducted in the last years to find the priority pollutants for management. One category of these screening procedure was developed to classify a inventory of a vast number of chemicals, which was hazard-based, and the other category of screening procedure categorized chemicals using a risk-based method, which considering the environmental concentration. It is a new trend to develop risk-based system in this field. The objective of this dissertation was to develop a risk-based screening approach for identifying and ranking the priority contaminants for risk-control. The first challenge of developing a risk-based screening procedure was the way to obtain the environmental concentration for a unidentified number of contaminants. High cost, tedious and time-consuming works prohibited the use of traditional quantitative analytical schemes for a large number of existing contaminants, due to the fact that most of them were designed for a specific class of contaminants with similar properties. In light of these considerations, a high throughout screening approach for identification and ranking of potential risky contaminants was proposed to overcome the drawbacks of traditional analytical methods designed for a specific class of contaminants. To this end, sample preparation was optimized to provide a high throughout extraction and cleanup for a broad range of contaminants. Based on the hypothesis that contaminants with similar function groups behaved similar chromatographic characteristics and responded in a comparable intensity on mass selective detector, semi-quantification of 928 organic contaminants was conducted by 15 analogous calibration standards. Under optimal conditions, the developed approach can be used to estimate concentrations of at least 1283 contaminants simultaneously with uncertainty lower than one order of magnitude and recoveries ranging from 38.6 to 128.6% on average. The high throughout screening approach was successfully applied to analyze the potential risk contaminants in sediments of Qingcaosha reservoir. The high throughout screening approach offered a database containing the quantification information of the identified contaminants for a site-specific area. Combined with existing toxicity data, a risk-based screening approach was established. The probability ecological risk assessment was integrated in the screening-level assessment to rank the potential risky organic contaminants to overcome the drawbacks of point estimate method used in risk assessment. The risk-based approach was validated to screening the priority contaminants for risk control in Haihe River. The results of different water systems were compared. The developed risk-based screening approach was applied to screen the priority contaminants in multi-medium, such as water, suspended particle and sediment in Liaohe River. Results showed that terbucarb was selected as the priority contaminants for risk control in the three medium. Moreover, 11 contaminants, including acetophenone, isophorone, o-phenylphenol, were listed as the priority pollutants in suspended particles. Meanwhile, 11 contaminants were selected as priorities in sediments. The comparison between results of our ranking to those from a more traditional score-based approach showed that a risk-based prioritization approach that incorporate concept of risks and estimate exposure using measured environmental concentrations is more reasonable for ranking contaminants posing risks to the environment. Results from this study indicated that the risk-based prioritization approach was feasible in site-specific screening analysis to identify and rank the unregulated contaminants. It would be useful in finding targets for environmental management. The developed risk-based approach in this study was feasible in screening risk assessements for different environmental multi-medium matrice.