Biological diversity provides many ecosystem functions and services that are critically important for human survival and development. In the past hundreds of years, biodiversity decreases continuously due to accelerating human activities and climate changes, leading to the alteration of ecosystem processes and ecosystems stability. An explicit understanding of risks assessment of biodiversity loss is essential for biological conservation. However, traditional risk assessment mainly focused on species diversity, which could not represent biological diversity comprehensively and reflect biodiversity loss at ecosystem level directly. The results of assessment can hardly be applied to policy-making for biodiversity conservation at landscape level. Furthermore, species-by-species is time consuming and resources costing. By the year 2010, less than 3% of the world's known species had been evaluated for potential inclusion in the International Union for Conservation of Nature (IUCN) Red List of Threatened Species. Hence, a higher-level biodiversity assessment may provide a more cost-effective means for multi-scale biodiversity conservation. Recently, more and more attentions have been tailored to develop a set of criteria for ecosystem risk assessment. Several protocols have been developed in many countries over the past twenty years, such as Finland, Germany, Denmark and Bulgaria. Most of these assessment protocols adopted the assessment system of the IUCN Red List of Threatened Species. Threat status was adopted as the major ecosystem risk assessment indicator, and threat status of each ecosystem was assigned using rule-based criteria based on thresholds for distributional and functional symptoms. However, there are significant differences between ecosystem classification, quantitative criteria, and spatial/temporal scale in different protocols, leading to possible distortions in results of assessment. In the year of 2008, IUCN created the Ecosystems Red List (RLE) Thematic Group that aims to develop a quantitative categories and criteria system, in order to establish a global standard for the ecosystem risks assessment that can be applied at local, regional, and global levels. The group released first version of ecosystem risk assessment system in 2011 which mainly consist of the definition and classification of ecosystems, identification of the threat status of ecosystems, quantitative criteria of ecosystem distributions and ecological functions, thresholds for these criteria, and standardized methods for assessments. IUCN Red List criteria for ecosystems version 2.0 was released two years later, along with twenty case studies to test the criteria. On the basis of previous version, the latest version introduced the criteria of environmental degradation and the ecosystem collapse, providing a workable and robust framework for ecosystem risk assessment. Up to now, there have been many countries that applied RLE categories and criteria to assess the ecosystem risk at regional or national scale, such as South Africa, Norway, Venezuela, Canada and New Zealand. Our study reviewed the ecosystem risk assessment protocols mentioned above, and found that IUCN Red List criteria for ecosystems are the most reasonable ecosystem risk assessment tools at present. We also briefly introduced the progress and future works of ecosystem assessment in China, and analyzed the importance and feasibility of applied RLE categories and criteria to assess ecosystem risk at national scale. We close by considering how to improve the ability of ecosystem management and biodiversity conservation using RLE and putting forward suggestions for developing RLE in China.