Interaural time differences (ITDs) are the dominant cues for human localisation of low-frequency sounds. Although a mechanism for ITD processing proposed in 1948 seems applicable to birds, and is consistent with many aspects of the responses found in mammals, recent data suggest that key tenets of the model might need to be reconsidered. The model requires, at every frequency, a distribution of cells with firing rate peaks across all ITD values within the animal's physiological range. The ITD tuning relies on internal delays in the form of a neural delay line. The evidence for such a delay line structure in mammals is not as convincing as it is in birds and, in some small animals the full range of physiological ITDs are not fully represented by peak firing of neurones at every frequency channel. Alternative means of achieving internal delays such as inhibitory inputs or the delays associated with cochlear filtering are being considered.