Urban governance of air pollution at the fine-grained or hyperlocal level has attracted great attention to improving the quality of urban life. The assessment of spatiotemporal patterns of fine particles (PM 2.5 ) in urban areas and their corresponding influencing factors can help reinforce the positive impact of urban planning and design on public health. Fine particulate matter (PM 2.5 ) was selected as the research object and the literature review of its distribution characteristics, influencing factors, the evolution of research methods was presented under the regional-city-community scale. The outlook on urban planning response was put forward then. As urban governance has entered into a new development stage in China, researchers have gradually started to explore the urban PM 2.5 distribution patterns in-depth in different spatial and temporal dimensions. In this way, they could respond to air pollution issues more efficiently and precisely from the perspective of urban planning and management. The annual variation of PM 2.5 concentration in China showed a U-shaped curve, which presented higher in winter and lower in summer. In addition, the daily concentration variations were bimodal and had hourly differences, showed that the maximum appeared during the morning and evening rush hours. The spatial distribution showed that the pollution caused by PM 2.5 was severer in the plain areas of North China, while the areas north of the Yangtze River also showed higher PM 2.5 concentration levels due to the dense population and developed economy. South of the Yangtze River, Tibet, and Yunnan in the southwest and southeast coastal areas had lower annual average PM 2.5 concentrations. At the urban scale, different cities showed different spatial and temporal distribution patterns due to their various geographical location, population distribution, and climatic characteristics. At the same time, the distribution of the pollution sources, traffic behavior, and blue-green spaces within the cities were also the main influence factor of PM 2.5 concentrations on the urban scale. The community-scale PM 2.5 distribution showed much heterogeneous, in specific, areas with dense traffic and crowd activities presented higher PM 2.5 concentrations, especially near highways. As urban management is required increasingly refined, understanding the complexity of the main factors affecting the multiple-scale spatial and temporal distribution of urban PM 2.5 , promoting refined urban air management according to local conditions, and carrying out the optimal assessment of urban planning and design based on air quality is becoming an inevitable option to achieve high-quality urban development. PM 2.5 concentration in cities is mainly governed by the land pattern, transportation network, emission source distribution, blue-green open space, and meteorological factors. At the regional scale, PM 2.5 concentrations were mainly related to economic and social factors, such as industrial structure, urbanization rate, and precipitation. At the city scale, urban spatial structure, land use type, ventilation corridor construction, population size, and meteorological conditions all affected the distribution of PM 2.5 concentrations in cities. In addition, the spatial layout of buildings and structures, the morphology of transportation facilities, and the characteristics of vegetation were identified as the main influencing factors of PM 2.5 distribution at the community scale. In terms of research methods, three types of methods were commonly used in studying the spatial and temporal characteristics of PM 2.5 concentrations in urban areas at multiple scales: atmospheric chemical transport models such as WRF-Chem and WRF-CMAQ, remote sensing inversion methods, and spatial statistical models such as land use regression (LUR). These methods could quantitatively assess and predict the spatial and temporal characteristics of PM 2.5 concentrations in high resolution. The development of artificial intelligence (AI) algorithms innovated the methods of urban air quality prediction simulation in temporal and spatial dimensions, providing more temporal and spatial data models for solving practical problems. They also promoted the formation of ever-complete spatiotemporal data prediction networks, which were more advantageous in the fusion analysis of nonlinear and more complex features at high precision scales. Using urban planning solutions, previous studies focused on two perspectives, reducing emission intensity and accelerating pollution dispersion to manage urban air pollution. The solutions of promoting regional collaborative governance and adjusting industrial structure were proved effectively in reducing pollution at the regional scale. Meantime, optimizing the plan of blue-green space and traffic facilities were useful solutions at the city scale. While at the community scale, adjusting the design of neighborhood building layouts was adopted. These solutions could help researchers deal with air pollution more efficiently and precisely in the field of urban planning and management. Finally, this paper proposed the following recommendations in response to the requirements of refined urban governance and existing urban governance experience: 1) Artificial intelligence (AI) models explore the PM 2.5 concentrations distribution characteristic and mapping of PM 2.5 source apportionment with high spatial and temporal resolution; 2) Determination of influencing factors of the spatiotemporal distribution of urban PM 2.5 concentrations under the different scales; 3) Prediction of high-resolution urban PM 2.5 concentrations distribution patterns and promotion refined urban planning and governance under different scales.