This paper proposes a framework model for the analysis of sound source localisation under the precedence effect using Gaussian mixtures. The model quantifies the effect of the lagging sound over the angular position and the width of the perceived auditory event, and incorporates some well-known aspects of the precedence effect in a single framework. Two sound source localisation experiments were carried out in order to show the utility of the proposed model. In the first experiment, the localisation accuracy for single sound sources was measured. In the second experiment, the subjects localised spectrally coherent lead-lag stimulus pairs with different angular separations. The responses were analysed using the proposed model. It is shown that the model can be used for assessing the fusion, localisation dominance, and the lag discrimination suppression aspects of the precedence effect as well as the associated widening of the auditory event.