A systematic numerical study was performed to investigate the influence of subgrid-scale (SGS) treatments on the simulation of turbulent free-jets. Large eddy simulations (LES) of such flows were performed to assess the accuracy of two SGS approaches: the detached eddy simulation (DES) and the dynamic Smagorinsky model (DSM). The numerical scheme based on the Non-oscillatory Forward-in-Time (NFT) and Multidimensional Positive Definite Advection Transport Algorithm (MPDATA) was employed to integrate the Navier-Stokes equations for incompressible flows. MPDATA due to its self-regularisation property is used to implicitly provide SGS effects. Two options of implicit LES (ILES) are investigated: ILES-NS, which solves the Navier-Stokes equations without an explicit SGS model, and ILES-EU that solves the Euler equations where viscous terms are absent. The performance of each approach was evaluated focusing on key global characteristics of jets, self-similar properties, and energy spectra. Quantitative and qualitative comparisons showed that all simulations were in good agreement with laboratory experiments, prior numerical studies, and each other, thus confirming the validity of the numerical approach and suitability of ILES for this class of flows. Additionally, energy spectra analysis revealed that ILES can reproduce the −5/3 and −7 gradients that signify the universal inertia subrange and dissipation range for turbulent free-jets.