Objective  Plant diversity is not only the basis of human survival and development, but also an important part of urban ecosystem. In view of the lack of evaluation standard of plant diversity for urbanization area, this study attempted to establish an evaluation system of plant diversity for Beijing urbanization area.

  Method  In this study, nine districts with a high degree of urbanization in Beijing were taken as the evaluation units. Through the collection of historical data and field investigation in key areas, the evaluation system of plant diversity in urbanization area of Beijing was constructed by six indicators, including richness of vascular plant species, diversity of ecosystem types, species endemism, degree of invasion of invasive alien species, rate of urban green space and richness of ancient and famous trees. The weight of each indicator was determined by regional diversity evaluation standard issued by the Ministry of Ecology and Environment, combined with expert opinions.

  Result  The plant diversity index of the nine districts in the study area was relatively high, 88.39 in Haidian District, 61.69 in Shijingshan District, 60.30 in Shunyi District, 49.75 in Fengtai District, 42.33 in Chaoyang District, 42.01 in Daxing District, 41.55 in Tongzhou District, 38.14 in Dongcheng and Xicheng districts; the evaluation grade of plant diversity was “very rich” in Haidian, Shijingshan and Shunyi districts, and “rich” in other districts.

  Conclusion  Taking district as basic unit, the evaluation system for plant diversity in urbanization area can reflect the overall status of plant diversity in Beijing. Evaluation result is the quantitative index for the assessment of ecological civilization construction in each district. Moreover, it can provide scientific basis for the conservation and management of biodiversity in Beijing. In the future, it is suggested that the regional grid (10 km × 10 km) should be used as the evaluation unit, and the indicators of the evaluation system should be continuously optimized. In this way, the evaluation results will be more fair and reliable, making the later protection measures more operable.