Despite the appearance of an enormous number of publications about urban ecology and species diversity, many issues are simply opened up rather than explained. The ecological complexity of urban areas, i.e., the variety of determinants and the spatial and temporal dynamic of cities, preclude simple starting points and lines of explanation. Therefore, we lack sufficient comparisons between various cities, especially comparisons on a global level. If cities are to be compared by appropriate indicators, and if they are to be evaluated with respect to urban biodiversity, then models are necessary that help us understand and mirror the causal relationships between urban areas and biological diversity. Three approaches, also representing a multiscaled view of urban areas, are presented that are suitable for developing applicable models and indicators for monitoring ecological systems: the embedded city, the urban matrix, and urban patches. The embedded city represents a globally useful concept, because the relationship between cities and their regions can be applied as an indicator to all regions. The lack of sufficient description of the urban matrix makes comparisons between cities difficult and causes scientists to underestimate the importance and function of the matrix for urban biodiversity. Urban patches are often investigated in urban studies about plants and animals. Therefore, much existing data can be used, and several criteria describing the functions of patches for biodiversity are available. In particular, the first two approaches should be researched more intensively.