Fires can be an important hazard for the safety of chemical and process industries. Particularly, pool fires are the most frequent fire scenarios in such facilities and can affect other equipment of the plant with severe consequences due to the domino effect. During the last decades, simplified fire modelling tools were used to predict some of the harmful effects that hydrocarbon pool fires may entail. Although these can be applied to limited number of scenarios, they cannot cover the overall characteristics governing the fire behaviour. Computational Fluid Dynamics (CFD) modelling can provide more detailed insights of the related fire effects, can consider complex geometries and can represent from small to large scale fires. However, simulation results should be firstly compared to experimental measurements in order to assess the predictive capabilities of these tools. This paper investigates the predictive capabilities of CFD modelling when performing a priori simulations of medium and large scale hydrocarbon pool fires. The main objective is to assess the fire effects prediction performance of two CFD codes that may be used to evaluate the hazard of hydrocarbon pool fires. FLACS-Fire and FDS codes have been used to simulate medium and large scale pool fires (1.5, 3, 4, 5 and 6 m-diameter) of diesel and gasoline fuels in unconfined environments. Given the notable differences between the mathematical methods applied to solve the CFD sub-models, the mesh resolution and the boundary conditions in each investigated tool, this study is not aimed at directly comparing both codes (i.e. using identical sub-models choices). However, the present CFD analysis is intended to reveal the potential of each software separately by applying the most appropriate modelling options for each tool. Based on a qualitative assessment of the predictions and a quantitative error estimation of the variables measured (i.e. flame temperature, burning rate, heat flux, flame height, flame surface, and surface emissive power), the main strengths and weaknesses of FLACS-Fire and FDS are identified when modelling hydrocarbon pool fires.