In this study, we assess the ability of two Volume Of Fluid (VOF) methods to simulate accurately emulsions in turbulent flows. Two-phase emulsions are prevalent in a wide spectrum of industrial applications. Pharmaceutical products, food processing, oil production and waste treatment are some of the areas where emulsions play an important role [1]. Since experimental studies of such flows are inherently difficult to conduct, we need to rely on numerical simulations in order to enhance the understanding of the underlying physics. Accurate interface capturing in such problems is fundamentally challenging due to the intrinsic, violent, change of the interface topology but also due to the multi-scale and intermittent nature of the flow. The dispersed phase undergoes continuous breakup and coalescence resulting in complex interface dynamics. The scope of this work is to investigate the differences in prediction accuracy between two popular interface-capturing methods. One is the MUlti-dimensionsal Limiter for Explicit Solu- tion (MULES) which has been extensively used in industrial applications [2]. The other is the geometric VOF method, isoAdvector, which has proved able to keep the interface sharp [3]. For this purpose, we perform Direct Numerical Simulations (DNS) of an emulsion of two im- miscible fluids with the same density and viscosity under the presence of forced tri-periodic Homogeneous Isotropic Turbulence (HIT). The DNS are carried out with the open-source, Com- putational Fluid Dynamics (CFD) toolbox OpenFOAM® [4]. We base the evaluation of the two methods on both algorithmic quantities (interface sharpness, volumetric conservation, bounding, computational cost) and physical quantities (Turbulent Kinetic Energy (TKE) spectra, Droplet Size Distribution (DSD), interfacial area evolution).