The present study is concerned with the projection of the lateral geniculate nucleus onto cortical area 18. Horseradish peroxidase (HRP) was injected into area 18 of 15 cats. Drawings were made to determine the location of the injection site and the distribution of labeled neurons in the lateral geniculate nuclei of each cat. The local retinotopic maps constructed prior to the injections and the reconstructions of the lateral geniculate nucleus were used to determine the location and the extent of each of the HRP injections. In 15 of the 25 hemispheres studied, the ratio of the number of HRP-labeled neurons in lamina A relative to the number of labeled neurons in lamina A1 was calculated. This ratio varied from 1.06 to 0.28, indicating that at least some regions of area 18 are dominated by inputs from lamina A1. However, if the HRP-labeled neurons in lamina C are included in the counts for lamina A, then the ratio A+C/A1 has a mean of 1.11, suggesting that area 18 receives a balanced input, with inputs from the contralateral eye being relayed through laminae A and C, and inputs from the ipsilateral eye being relayed through lamina A1. When the distribution of HRP-labeled neurons in lamina A was plotted onto a dorsal view of the lateral geniculate nucleus, the labeled neurons formed an ellipse with the long axis of the ellipse oriented parallel to the isoelevation lines. The representation of azimuth is compressed in area 18 relative to the lateral geniculate nucleus. In six hemispheres the injections were restricted to a few layers of the area 18. Following small injections into layer IV of area 18, the HRP-labeled neurons occupied an extensive region of the lateral geniculate nucleus, indicating a considerable amount of convergence of the inputs to area 18. In hemispheres where the injections were restricted to layers I and II, labeled neurons were only seen in the medial interlaminar nucleus and the C laminae.