The intensive three-dimensional urban landscape has become a trend in urban mode development, which might be one of the effective ways of solving problems such as the shortage of living space in the city, and the contradiction between three-dimensional urban elements and the urban bases. The three-dimensional city would become one of the important city models in the future. For urban landscape ecology, the spatial effect of configurations and functions in urban landscape not only exists in the plane space, just as the maps showed, but also affects the spatiotemporal change of urban landscape in three-dimensional space. As a result, study on the spatial pattern and functional characteristics should not be localized in the two-dimensional plane. This paper reviews the concepts, research method and experience of two-dimensional urban landscape, and advances the basic characteristics of three-dimensional urban landscape. The results show that three-dimensional landscape pattern and landscape functional network are the basic important characteristics of three-dimensional urban landscape. First of all, retrieval of three-dimensional information was the precondition for studying three-dimensional urban landscape. Currently, a series of techniques have been successfully used to retrieve the variation of urban landscape in the vertical scale from remotely-sensed data, including Light Detection and Ranging (LiDAR), Interferometric Synthetic Aperture Radar (InSAR) and Airborne Photogrammetry, etc. The paper introduces in detail the basic steps for the retrieval of vertical characteristics of building, road, tree and shrub from LiDAR data. Three-dimensional information extraction from LiDAR data could be simplified as the processes of subtraction of DTM from original DSM, which would generate the so-called normalized DSM (nDSM). nDSM could be seen as the representation of objects rising from the terrain placed approximately on a plane. Conventional urban landscape pattern analysis means the integrated results of different landscape types, including composition and configuration. When urban landscape vertical heterogeneity is considered, decomposition of urban landscape into various components with different vertical characteristics using remotely sensed data is a new research field which could provide more economical and ecological meaningful parameters for quantifying the urban environment and its changes. First, this paper reviews in detail the latest achievements and frontiers of three-dimensional urban landscape ecology. Secondly, by comparing and analyzing the main methods and models in the retrieval of three-dimensional landscape elements, we propose a research framework for three-dimensional urban landscape ecology, including the extraction of three-dimensional urban landscape elements, measurement of three-dimensional landscape pattern characteristics, detection and modeling of three-dimensional landscape dynamics, ecological and environmental effects on biodiversity and local climate change, and urban landscape-function relations based on the concept of a landscape functional network. Finally, this paper discusses the existing problems and gives an outlook of future research directions.