Introduction. We used high-field (3T) functional magnetic resonance imaging (fMRI) to map the retinotopic organization of human cortical areas. Methods. Retinotopic maps were reconstructed using existing mapping techniques. Stimuli were made of a rotating wedge stimulus, which provided angular coordinate maps, and an expanding or contracting ring, which provided eccentricity coordinate maps. Stimuli consisted of a grey background alternating with a flickering checkerboard. A Brucker 3T scanner equipped with a head coil and a custom optical system was used to acquire sets of echoplanar images of 20 occipital coronal slices within a RT of 2.111 ms and an 8 mm(3) voxel resolution. Surface models of each subject's occipital lobes were constructed using the Brainvisa software from a sagittal T1 -weighted image with a 1 mm(3) voxel resolution. The cortical models were then inflated to obtain unfolded surfaces. Statistical analyses of the functional data were made under SPM99, and the response amplitudes were finally assigned to the cortical reconstructed surfaces. Results. We identified boundaries between different early visual areas (V1, V2, V3) using eccentricity and polar angle retinotopic maps and detection of reversals in the representation of the polar angle. Both complete maps and reversals corresponding to the V1/V2 borders were clearly visible with a single recording session. Also, we were able to compare data from subjects across various fMRI acquisitions and found that there was a strong correlation between maps acquired at different sessions for the same subject. Conclusions: We developed a quick (<40 min) retinotopic cortical area mapping method at 3T, which makes it possible to study the cortical remapping in patients with retinal scotomas.