The afferent connections of the primate's temporopolar cortex were investigated with the retrograde horseradish peroxidase technique. Old World and New World monkeys received small unilateral injections of horseradish peroxidase. These labeled cells in a number of cortical, thalamic, and brainstem regions and in a few further telencephalic and diencephalic regions. Cortically, the neighboring areas of the inferior and superior temporal gyrus and the insula contained a considerable number of labeled cells. Furthermore, a substantial projection arose from the orbitofrontal and the frontopolar cortex. The cingulate gyrus contained only very few labeled cells. Interhemispherically, corticocortical connections arose mainly from temporal lobe areas. Labeled cells were seen in various regions of the basal forebrain and cells labeled only faintly in the lateral and basal accessory nuclei of the amygdala. The claustrum contained labeled neurons only in one rhesus monkey. On the diencephalic level, the caudal medial portion of the medial pulvinar was the principal thalamic source of afferents to the temporopolar cortex. Furthermore, labeled cells were found in the neighboring, caudal part of the mediodorsal nucleus, within and along the nucleus limitans, in the medial geniculate nucleus, and in several nuclei of the nonspecific system. The fields of Forel, the zona incerta, and lateral and dorsomedial hypothalamic areas contained a few labeled cells. Within the brainstem of the rhesus monkeys those regions projecting diffusely to the cortex contained a few labeled neurons. Furthermore, these brains had some labeled cells in the regions of the nuclei medialis annuli aqueductus, tractus mesencephalicus nervi trigemini, and trochlearis. Although among the three species differences in the cortical and thalamic projection patterns were observed, the regions projecting most densely to the temporal pole were similar in principle. This statement holds in particular for cortical and thalamic sites. However, the greatest number of labeled cells was found in the rhesus monkey, a fact that cannot be attributed solely to the size of the horseradish peroxidase injections and the size of the brain, but that appears rather to represent a true species difference. From our results we conclude that the temporopolar cortex constitutes a cortical area necessary for effective affectional-sensory integration.