The accessibility and equity of green spaces are important indicators to measure the environmental, ecological and social benefits of urban green spaces, and also important manifestations to evaluate and reflect the scientific and rational spatial allocation of urban green spaces resources, which are closely related to urban environment, ecology, health, society and sustainable development. Previous studies seldom measure the accessibility and equity of urban green spaces from multiple scales and multiple travel modes. In addition, the traditional method of ArcGIS network analysis is mostly used to estimate the green spaces accessibility of a single travel mode (such as car or walking), while the green spaces accessibility of different travel modes is seldom considered and measured at the same time, and different people have different choices for travel modes. It is necessary to explore the measurement method of green spaces accessibility considering multiple travel modes, so as to provide a scientific basis for a more comprehensive study of the rationality, accessibility and equity of the spatial distribution of urban green spaces resources. In this paper, taking Guangzhou as the research area, combined with the Travel O-D Point Intelligent Query System, a Multi-mode Two-step Floating Catchment Area Method Modes and a green spaces equity model are constructed respectively, and the spatial pattern and difference of green spaces accessibility and equity in Guangzhou are studied. The results show that (1) there are distinct "suburban polarization" spatial differences in the accessibility level of multiscale green space in Guangzhou. Areas of low green spaces accessibility are concentrated in the central city with higher population density, while the accessibility value of green spaces in the periphery with a smaller population is generally higher. (2) The spatial distribution of green spaces in Guangzhou is seriously unfair (the green equity index is 0.58). There are also significant differences among administrative regions. Nansha District, Tianhe District, and Huadu District have good fairness in green spaces, and the spatial distribution of green space in other regions needs to be improved and optimized. (3) At the subdistrict level, 73.29% of the green spaces are distributed at a high average level (green spaces equity index≤0.2). At the community level, the number of communities with low accessibility-equity type is the largest (accounting for 39.85% of the total community). The communities with high accessibility type are the least distributed: 15.91% of the communities belonging to high accessibility-equity type and 12.94% of the communities belonging to high accessibility-inequity type. The research conclusion will provide scientific bases for optimizing the spatial structure of urban green spaces and ensuring the equity of green spaces.