Abstract The understanding of processes that govern soot production in aero-engines is fundamental for the design of new combustion systems with low environmental impact. For this, the development of numerical models that allow for soot accurate prediction for a limited computational cost is required. Many combustors, more specifically those used in aero-engines, feature rich flame regions typically exploited in the so-called Rich-Quench-Lean (RQL) technology. Thus, it is important to complete the data on soot formation in turbulent flames that exist in the literature, by considering rich turbulent flames operating in the premixed mode. To this purpose, a model scale swirled combustor, called EM2Soot, was designed at the EM2C laboratory to analyze soot production under perfectly premixed rich conditions. In this work, the effect of the equivalence ratio on soot production in this burner is experimentally characterized and numerically simulated. Measurements of Planar Laser Induced Fluorescence of Polycyclic Aromatic Hydrocarbons (PLIF-PAH) were performed to examine soot precursors presence, whereas soot volume fraction is measured with Planar Laser Induced Incandescence (LII). Large Eddy Simulations (LES) are carried out using an analytically-reduced chemistry for the gas phase simulation in combination with a three-equation model for the solid phase description. By considering a range of equivalence ratios, the soot volume fraction from the experiments was found to reach a maximum near ϕ = 1.8, whereas a lower level of soot volume fraction was measured for lower and for higher equivalence ratios. The large eddy simulations performed in parallel are found to be in qualitative agreement with experimental data in terms of PAHs location and soot presence. The soot volume fractions fv are notably overestimated with respect to the LII measurements, suggesting the need for improved modeling schemes. However, the numerical results correctly retrieve a reduction of soot production for the highest considered equivalence ratio value and can be used to explain the experimental behaviour. Results indicate that LES gives access to a qualitative understanding of the interactions between flame, turbulent flow, and precursors governing soot formation in this system.