This study presents a methodology to predict particle number (PN) generation on a naturally aspirated 4-cylinder gasoline engine with port fuel injection (PFI) from wall wetting, employing numerical CFD simulation and fuel film analysis. Various engine parameters concerning spray pattern, injection timing, intake valve timing, as well as engine load/speed were varied and their impact on wall film and PN was evaluated. The engine, which was driven at wide open throttle (WOT), was equipped with soot particle sampling technology and optical access to the combustion chamber of cylinder 1 in order to visualise non-premixed combustion. High-speed imaging revealed a notable presence of diffusion flames, which were typically initiated between the valve seats and cylinder head. Their size was found to match qualitatively with particulate number measurements. A validated CFD model was employed to simulate spray propagation, film transport and droplet impingement. Wall film properties were carefully analysed on the nearby surfaces were diffusion flames were spotted.Results showed a good correlation between the PN and the amount of wall film mass on the junction between the cylinder-head and intake ports. Major in-cylinder liquid transport mechanisms, which were found to influence wall wetting on the cylinder head-intake ports regions were clarified. Finally, the impact of gas-exchange, injection timing and spray pattern on film transport mechanisms and wall wetting was assessed.