While current assessment methods for preventing progressive collapse are mainly associated with blast and impact loading, no systematic framework is currently available for the practical and rational assessment of robustness of multi-storey building structures under localised fire. In this thesis, a robustness assessment framework with various alternative levels of sophistication is presented with the aim of bridging the gap between the topics of structural fire resistance and progressive collapse. The robustness assessment framework developed in this thesis is comprised of four basic components, namely, detailed Temperature-Dependent Approach, simplified Temperature-Dependent Approach, Temperature-Independent Approach, and practical design recommendations. These assessment approaches can satisfy various design requirements in different design stages. To illustrate their application, localised fire induced by burning vehicles in a typical multi-storey steel-composite car park is considered as a main reference scenario. A Robustness Limit State (RLS) is proposed, which is based on the fact that large inelastic deformations of building structures subject to extreme loading are typically concentrated in the joint regions, thus failure of joints in certain locations may lead to floor collapse, and subsequently trigger progressive collapse. Therefore, no floor collapse is allowed in the current RLS. Since joint resistance and ductility play an essential role in mitigating progressive collapse, a component-based joint modelling technique and multiple joint failure criteria for commonly used semi-rigid joints are proposed and thoroughly discussed. The proposed joint modelling strategy is shown to be capable of capturing realistic nonlinear behaviour of joints under both ambient and elevated temperatures. Employing the proposed joint failure criteria, the application of the robustness assessment framework is illustrated for the reference structure, and important conclusions are drawn relating to the accuracy and reliability of each assessment approach. Furthermore, multi-level structural modelling strategies, the significance of structural dynamic effects during fire, factors influencing structural robustness, and future research