This research aims to assess vulnerable spaces around heritage buildings concerning their socio-spatial properties. Additionally, the research explores the predictive relationship between these properties and contextual anthropogenic hazards. The research's methodology relies on multi-methods applied to twenty-eight heritage buildings in historic Cairo, Egypt. Firstly, the research employed the Delphi technique and ICCROM-CCI-RCE method to assess the potential rates of contextual anthropogenic hazards in the study area. Afterwards, the literature review was conducted to explore a new paradigm for assessing vulnerable spaces using the space syntax-based methodology. Space syntax provides a better understanding of space, its structure, and how it affects human behaviour. Moreover, the research employed two main analytical methods of space syntax, axial graph, and visual graph analyses, which were achieved by “Depthmap 4” software to investigate the syntactic context of the study area. Furthermore, the exploratory factor analysis was employed to statistically analyse the syntactic output data to develop fewer factors of socio-spatial vulnerability. These factors served as inputs for multiple linear regression analysis as predictive models of the influence of socio-spatial vulnerability on the assessed contextual anthropogenic hazards. Finally, the resulting models highlighted the importance of investigating the socio-spatial properties around heritage buildings to predict human destructive behaviours based on quantitative analytical methods. Such results would help authorities formulate suitable and sustainable strategies for the adequate performance of heritage buildings. Also, the predictive models can potentially be used in other livable historic cities.