Transneuronal autoradiography was used to study the role of visual experience in the development of ocular dominance patches in the cat. In order to assess quantitatively the effects of visual deprivation, image analysis was used to measure the profiles of grain density in layer IV. Fourier power spectra of these profiles were computed to give a measure of the amplitudes and frequencies of the fluctuations in grain density that were present. Deprivation of normal patterned vision by binocular lid suture or by rearing in total darkness from shortly after birth abolished the dominant periodicity (of about 1.1 mm) in the distribution of left and right eye afferents in layer IV of area 17. A dominant periodicity of about 2.2 mm was, however, present in area 18 of both normal and dark-reared animals. Visual deprivation was not able to reverse segregation. One animal reared normally for 6 weeks was placed in the dark for a further 28 weeks and showed normal periodicities in the distribution of geniculate inputs to area 17. Another animal given 128 hours of experience and kept in the dark for the rest of the time until it was 12 weeks old also showed normal segregation. To determine the minimum amount of visual experience necessary for segregation to occur, four animals were given 8-, 24-, 48-, and 128-hour periods of visual experience and were studied at 12 weeks of age. Eight hours of experience had no detectable effect on segregation; periodicities of intermediate amplitude were present in animals that received 24 and 48 hours of experience, while 128 hours of experience resulted in periodicities of normal amplitude. Recovery from visual deprivation was studied by rearing kittens from birth in the dark for varying periods and then returning them to the normally lit colony room for periods of 6 to 22 weeks. Recovery from 6 weeks of dark rearing was found to be complete; much less recovery occurred following periods of 8 to 25 weeks of initial deprivation, and no recovery at all occurred after 30 weeks of deprivation. It is concluded that the spontaneous activity present in the geniculocortical afferents of dark-reared and lid-sutured cats is not adequate to drive normal periodic segregation in area 17, though it can do so in area 18. Between 48 and 128 hours of visual experience, given before 8 weeks of age, appears to be necessary and sufficient for normal periodic segregation of geniculate afferents in area 17 of the cat.