This thesis gives an overview of my work over the last four years on the development of analogue electronic building blocks for the auditory pathway, and their application to some models of processing in the auditory brainstem. The anatomy and physiology of the human ear is presented, and is decomposed into three key elements, i.e., the basilar membrane 'band-pass' filters, the transduction into a neural signal performed by the inner hair cells, and the mechanical feedback introduced by the outer hair cells. An electronic model for the first two of these elements is presented and measurement results are shown to compare these circuits with their biological counterparts. The remaining part of the human auditory pathway consists of several groups of different types of spiking neurons. Since the main part of signal processing in the auditory pathway is performed by these different types of spiking neurons, a good spiking neuron model is essential. The electrophysiology and anatomy needed to understand the basics of the spiking behaviour of these neurons is presented. If we want to model large groups of spiking neurons, the neuron circuit also needs to be small. I propose a circuit that is simple and small, yet capable of emulating several types of neurons found in the Cochlear Nucleus, as shown by chip measurements. With these electronic building blocks I can start to model neural architectures in the brain that extract certain signal characteristics, and these models will operate in real time. Although only few of such architectures have been identified to date, once the types of neurons in these brain circuits. and their inter-connectivity is known, it is then fairly straightforward to create an analogue VLSI model. I present two examples, both based on periodicity detection, since this is a domain where spike based operation seems especially useful. The first example uses synchronized activity on auditory nerve fibres from two positions along the basilar membrane to obtain a high frequency selectivity and a representation of the sound which is independent of intensity. This model is completely hypothetical, since no evidence has been found in the brain for its existence, but it provides an excellent introduction into the periodicity detecting power of spike based computation. The second example is a model of amplitude modulation sensitivity in the inferior colliculus. This model can be used for example to extract the modulation frequency of an amplitude modulated sound, or to extract the fundamental frequency of a harmonic complex. These sounds are of special interest because many natural sounds such as animal calls or speech fail into this category. This model has a much stronger biological basis, since neurons have been found in the inferior colliculus that react to amplitude modulation in a similar way as the model.