Human area 17 is known to contain a single (the primary) visual area, whereas areas 18 and 19 are believed to contain multiple visual areas (defined as individual representations of the contralateral visual hemifield). This is known to be the case in monkeys, where several boundaries between visual areas are characterized by bands of callosal afferents and/or by changes in myeloarchitecture. We here describe the pattern of callosal afferents in (human) areas 17, 18, and 19 as well as their cortical architecture and we infer the position of some visual areas. Sections from occipital lobes of 6 human brains with unilateral occipital infarctions have been silver-impregnated for degenerating axons, thereby revealing callosal afferents to the intact occipital cortex. Their tangential distribution is discontinuous, even in cases with large lesions. A band of callosal afferents straddles the area 17/18 boundary, whereas the remainder of area 17 and a 15-45 mm wide stripe of area 18 adjacent to the callosal band along the 17/18 border are free of them. Patches of callosal afferents alternate with callosal-free regions more laterally in area 18 and in area 19. We conclude that, in man, a second visual area (analogue of V2) lies in area 18, horseshoe-shaped around area 17, and includes the inner part of the acallosal stripe adjacent to the callosal band along the 17/18 boundary. The outer part of this acallosal stripe belongs to a third visual area, which may contain dorsally the analogue of V3 and ventrally that of VP. Thus the lower parts of the second and third visual areas lie on the lingual gyrus, whereas the analogue of the macaque's fourth visual area probably lies on the fusiform gyrus. Although the proposed subdivision of the occipital cortex relies largely on the pattern of callosal afferents, some putative human visual areas appear to have distinct architectonic features. The analogue of V2 is rather heavily myelinated and its layer III contains large pyramidal neurons. Its upper part is not well delimited laterally since adjacent "V" has similar architecture. Its lower part, however, differs clearly from the adjacent "VP," which is lightly myelinated and lacks the large pyramids in layer III. The cortex lateral to "VP" is heavily myelinated and contains fairly large pyramids in layers III and V. The myeloarchitecture of the lateral part of the occipital cortex is not uniform; a very heavily myelinated region stands out in the lateral part of area 19, near the occipito-temporal junction.(ABSTRACT TRUNCATED AT 400 WORDS)