This article presents a numerical study of the fuel and turbulence distributions in a pre-chamber at spark-time. The study has been conducted in the framework of the H2020 Gas-On project, dealing with the development of a lean-burn concept for an automotive-sized gas engine equipped with a scavenged pre-chamber. The test case considered studies a 7-hole pre-chamber with circumferentially-tilted orifices mounted on the cylinder head of a rapid compression-expansion machine (RCEM), consistent with the experimental test rig installed at ETH Zurich. An accurate description of turbulence and fuel distributions are key quantities determining the early flame development within the pre-chamber. Both quantities have an influence on the overall combustion characteristics and therefore on the engine performance. For this purpose, computational fluid dynamics (CFD) is employed to complement experimental investigations in terms of data completeness. The performance of the Reynolds-averaged Navier-Stokes (RANS)-based turbulence model is compared with large-eddy simulation (LES) through ensemble averaging of multiple LES realizations, in which the fuel injection rate evolution into the pre-chamber has been perturbed. Overall, RANS results show that the distributions of the turbulent kinetic energy and fuel concentration at spark-time agree well with the LES ensemble-averaged counterparts. This constitutes a prerequisite in view of the combustion phase and the accuracy reported provides further confidence in this regard.