Today, ever changing and advancing techniques of construction are constantly pushing the envelope of structural possibilities in the built environment. Although not new, the concept of Double-Skin Façades (DSF) finds increasing implementation with the advent of sustainable construction, aiming to reduce energy consumption to condition buildings whilst improving indoor air quality. As is the case with the traditional concept of the compartment fire, methodologies and assumptions on which our general understanding of the fire problem is based, did fundamentally not change. Inherently bound to this, is the concept of compartmentalisation, prescribing measures to avoid horizontal and vertical fire spread in buildings. A DSF, most commonly featuring a ventilated cavity between curtain wall and the secondary glass façade at an offset, is prone to drastically alter fire and smoke behaviour once able to enter. Unlike curtain walls, the chimney-like aspect ratio of such façades is able to trap fire and combustion gases within the cavity, potentially compromising the integrity of the building perimeter above the fire. The current approach to this issue tends to focus on using non-combustible construction materials and the installation of sprinkler systems to avoid breakage of window panes in the first place. Another topic of interest is the weak connection between floor slab and curtain wall which can allow vertical fire spread to adjacent floors. Research has also been discussing the use of mullions to deflect the fire plume away from the façade. Even if useful in DSF’s, aesthetics and problems with functionality will most likely prevent mullions from being introduced into the DSF. However, very little relevant research actually investigated the fire-induced flow structure under these conditions so that properly informed design decisions can be made. The project at hand aims to understand hazards to the floors above and below the fire floor by experimentally investigating the governing processes by means of large-scale fire testing and small-scale salt-water modelling (SWM). The gathered data shall serve as a basis to discuss current spandrel and cavity design decisions. Results have been compared in terms of dimensionless numbers and demonstrate complex interactions between DSF cavity width and spandrel height, encouraging a discussion about the need of further research of this topic.