It is known from previous studies that fibers originating from cells in area 2 of the cat primary somatosensory cortex project topographically to area 4 of the motor cortex and that they terminate preferentially in the caudal region of the cruciate sulcus. We examined this pathway to determine more precisely the distribution pattern of fibers and the laminar arrangement of axon terminals in the motor cortex. The recently developed technique of PHA-L staining enabled us to label anterogradely the axons that form this projection. Iontophoretic injections of PHA-L were made into the rostral bank of the ansate sulcus (area 2). After 7 days the cats were perfused and the tissue was processed immunohistochemically to stain the PHA-L filled fibers. Light microscopic examination revealed that a small cluster of cells in the sensory cortex gave rise to multiple foci of labeled axons in area 4. The labeled fibers formed columnlike arrays, which were located for the most part in the posterior bank of the cruciate sulcus and were separated by irregular intervals of cortex devoid of labeled fibers. Clusters of labeled fibers were also found in the anterior region of the cruciate sulcus in some of the animals. The dimensions of the labeled areas and the small number of cells that gave rise to each group of fibers suggested that axonal branches of cells within the injection site formed the multiple foci. Variations in the immunohistochemical staining enabled us to study the laminar distribution of sensory cortex axon terminals with the electron microscope. Whereas some PHA-L labeled terminals were found in the deep cortical layers, the majority (82%) were spread throughout layers I-III. Differences in the laminar distribution of sensory cortex afferents that formed axodendritic or axospinous synapses were noted. Synapses formed with dendritic shafts were relatively sparse (28%) and were confined to the superficial layers. Some of the more numerous axospinous synapses, which accounted for 72% of identified synapses, were found in layers V and VI, although most were in layers I-III. The distribution pattern of terminals showed little variation between columns in different areas of the motor cortex, including that in the anterior cruciate region. The pattern of termination of the sensory to motor cortex projection is discussed in relation to the physiological characteristics of this pathway.