Different cortical and subcortical structures present neurons able to integrate stimuli of different sensory modalities. Among the others, one of the most investigated integrative regions is the Superior Colliculus (SC), a midbrain structure whose aim is to guide attentive behaviour and motor responses toward external events. Despite the large amount of experimental data in the literature, the neural mechanisms underlying the SC response are not completely understood. Moreover, recent data indicate that multisensory integration ability is the result of maturation after birth, depending on sensory experience. Mathematical models and computer simulations can be of value to investigate and clarify these phenomena. In the last few years, several models have been implemented to shed light on these mechanisms and to gain a deeper comprehension of the SC capabilities. Here, a neural network model (Cuppini et al., 2010) is extensively discussed. The model considers visual-auditory interaction, and is able to reproduce and explain the main physiological features of multisensory integration in SC neurons, and their acquisition during postnatal life. To reproduce a neonatal condition, the model assumes that during early life: 1) cortical-SC synapses are present but not active; 2) in this phase, responses are driven by non-cortical inputs with very large receptive fields (RFs) and little spatial tuning; 3) a slight spatial preference for the visual inputs is present. Sensory experience is modeled by a “training phase” in which the network is repeatedly exposed to modality-specific and cross-modal stimuli at different locations. As results, Cortical-SC synapses are crafted during this period thanks to the Hebbian rules of potentiation and depression, RFs are reduced in size, and neurons exhibit integrative capabilities to cross-modal stimuli, such as multisensory enhancement, inverse effectiveness, and multisensory depression. The utility of the modelling approach relies on several aspects: i) By postulating plausible biological mechanisms to complement those that are already known, the model provides a basis for understanding how SC neurons are capable of engaging in this remarkable process. ii) The model generates testable predictions that can guide future experiments in order to validate, reject, or modify these main assumptions. iii) The model may help the interpretation of behavioural and psychophysical responses in terms of neural activity and synaptic connections.