Urban studies have gradually expanded their vision from horizontal to volumetric dimensions, responding to the growing compactness and density of cities. Though land use study has a long history, the study of urban space use from a volumetric perspective remains limited. However, vertical spatial difference is as substantial as horizontal spatial variance, especially in mixed-use developments. Addressing this, the research analyses 3D floor space use dynamics in the City of Sydney between 2007, 2012, and 2017 using ArcGIS Pro and visualising results with Plotly. Utilising Voxel Automata and Markov transition logic in Netlogo3D, we simulate potential future 3D urban structures. The model transforms floor spaces into voxels, assigns varying transition probabilities to voxels based on self-state, and applies the influence of neighbourhood state. The research underscores the challenges in developing varied transition probabilities for different floors, revealing the complexity of modelling 3D space use dynamics. The findings provide a more realistic understanding of the complex urban system and cities’ volumetric development. Additionally, the utilised 3D visualisation method can extend its utility beyond floor usage types to other spatial variables. Consequently, the research highlights the importance of 3D system thinking in future urban growth and expansion studies, and suggests more precise transition rules for modelling specific time points, benefiting future smart governance.