Pre-fermentation of influent has been proposed as a method to improve phosphorus removal capacity, granulation, and simultaneous nitrification denitrification for aerobic granular sludge plants. This thesis sought to determine the influence of increasing VFA concentrations on performance of aerobic granular sludge plants and key considerations for implementation of pre-fermentation. Two approaches were applied during this thesis: a modelling approach and an experimental approach. A mathematical model was built to show the influence of increasing VFA concentrations and to describe principles observed in practice at aerobic granular sludge plants. The experimental approach consisted of two measuring campaigns. First, experiments were performed to test the necessary operational conditions to ferment influent in a sequencing batch reactor fermenter. Second, a sampling campaign was conducted to determine the conditions which led to fermentation of organic matter in a sand trap on site in Utrecht. The model demonstrated the benefit of combining existing models to describe settling of aerobic granular sludge with size fractionations, plug flow feeding in a settled sludge bed, reactions in the bulk and in granules, and diffusion across granules and into bulk liquid. It was found that increasing acetate concentrations led to longer stability of phosphorus uptake capacity and that SRT control is necessary to maintain long term capacity. Increasing acetate concentrations also led to increases in phosphate release and simultaneous nitrification denitrification. These changes were attributable to differences observed across granules of different sizes: phosphate accumulating organisms dominated larger granules and nitrifiers were more dominant in smaller granules. Model calculations indicate that a maximum granule size exists because the most poly-phosphate storage was not performed by the largest granules. From the experimental work, fermentation was found to be occurring (at rates of 21 mg VFA-COD/L/h ± 15) in a hydraulically underloaded sand trap which led to significant VFA production and subsequent difficulties in fermenting in the controlled SBR reactor (which produced an average of 28 mg VFA-COD/l). More research is needed to determine the optimal operational conditions to ferment influent and to demonstrate the effect on aerobic granular sludge in practice.