Both anterograde (autoradiographic) and retrograde (horseradish peroxidase) tracing techniques were used to identify and characterize projections from the dorsal lateral geniculate nucleus (DLG) of the thalamus to the subdivisions of occipital prestriate cortex (generally defined by areas 18 and 19) in macaque monkeys. It was found that the DLG projects to area 19 and the anterior portion of area 18 located on the lateral, ventral, and ventromedial surfaces of the hemisphere. There is a general topographical organization such that the medial portion of the DLG projects to dorsal prestriate located between the lunate and superior temporal sulci, while the lateral portion of the DLG projects to ventral prestriate extending from the inferior occipital to the occipitotemporal sulci. The DLG projects to most of the occipitotemporal sulcus. In contrast DLG inputs to the lunate, superior temporal, and inferior occipital sulci are limited in extent, and involve only a portion of the bank or shoulder of each sulcus which is continuous with the cortex located on the surfaces of the surrounding preoccipital gyri. The projection to the inferior occipital sulcus is more extensive than the ones to the lunate and superior temporal sulci and involves the floor as well as the ventral bank. This means that there are certain functional subdivisions, such as those located within the lunate (eg, visual area 2) and superior temporal sulci which do not receive DLG input. Regardless of the location of these projections the terminal pattern was the same and occurred in horizontally segregated "patches" which were restricted to layer V and the lower, adjacent portion of layer IV. In contrast, the projections from the nearby pulvinar to cortical layers IV, III, and I of the same prestriate areas do not overlap with those from the DLG. An attempt was made to identify possible afferent inputs to these DLG-prestriate paths. The HRP experiments reveal that the DLG-prestriate cells are concentrated in the DLG interlaminar zones and that their distribution overlaps the distribution of terminals from the superior colliculus and, as shown previously, prestriate cortex. Intraocular injections of radioactive precursors demonstrated transsynaptic transport to a number of structures except prestriate cortex. While the latter result does not prove that the DLG-prestriate cells do not receive retinal input, one conclusion is that the DLG-prestriate projection could be organized like pulvino-prestriate systems which receive their inputs from the midbrain and cortex.