Despite extensive behavioural work on the rat superior colliculus, its descending efferent pathways have not been fully characterised with modern anatomical tract-tracing techniques. To investigate these pathways, wheatgerm-agglutinin conjugated with horseradish peroxidase (1%) was injected at various locations within the superior colliculus of hooded rats. Label judged to be transported orthogradely was plotted on coronal sections modified from the atlas of Paxinos and Watson (1982). Two major descending pathways were identified. (i) The bulk of the fibres in the ipsilateral descending pathway leave the superior colliculus ventrolaterally, and course around the lateral margin of the midbrain reticular formation. Caudally, projecting fibres leave the main bundle to innervate the cuneiform nucleus, and parts of the pontomedullary reticular formation. Terminal fields associated with the major bundle of fibres are found in an area medial to the brachium of the inferior colliculus; the parabigeminal nucleus and adjacent tegmentum; the ventrolateral midbrain reticular formation; and the lateral pontine nuclei. (ii) The fibres of the main contralateral descending pathway leave the superior colliculus ventromedially, to cross midline in the dorsal tegmental decussation. They immediately turn caudally to join the predorsal bundle, in which they run the length of the brainstem to reach the cervical spinal cord. Major terminal fields occur in nucleus reticularis tegmenti pontis; the pedunculopontine/parabrachial area; paramedian pontomedullary reticular formation; and inferior olive. In addition there is lighter labelling in many areas of the pontomedullary reticular formation and in the cervical spinal cord. There was also a much sparser contralateral descending projection that crossed midline in the tectal commissure, and sent terminals to the contralateral cuneiform area and adjoining regions. These results suggest that the distribution of the descending efferent pathways from the superior colliculus in rats is similar to those described in other species. The fact that the two major pathways project to quite different terminal areas, together with previous findings that they have separate cells of origin within the tectum, suggests that they may also be functionally distinct.