Mine ventilation models are now an accepted component of most mine management systems. However, modern ventilation models also represent a highly structured spatial data source, defining underground pathways and connections across a mine. This creates opportunities to extend the use of network models beyond ventilation design, by using critical path algorithms to find pathways such as minimum distance or minimum time.This paper explores the use of ventilation network models to automatically analyse and calculate potential emergency escape routes. A variation of Dijkstra's algorithm is used to predict complex pathways between any two points across a model, representing an escape route from an emergency area (such as a fire) to a safe location (such as the surface or a refuge station). Weighting can be assigned to paths to represent potential travel speed, obstructions or preferred routes. Finally, emergency considerations such as smoke or gas in some parts of the mine can then be incorporated into the pathways to force the algorithm to avoid dangerous routes and specify safer pathways. Ventsim is used to visually show example results.Emergency egress pathway prediction using ventilation network models can provide an efficient way to quickly predict the quickest or most efficient escape routes for emergencies or planning purposes.CITATION:Stewart, C M, Aminossadati, S M and Kizil, M S, 2017. Emergency egress pathway prediction using ventilation models, in Proceedings Australian Mine Vent Conference 2017, pp 57-62 (The Australasian Institute of Mining and Metallurgy: Melbourne).