1. The activity of single units was recorded from the striate cortex (area 17) of anaesthetized, paralysed cats. Responses to stimuli moving at different velocities were examined. 2. Peak evoked firing frequency, rather than fotal evoked spikes, is used throughout as a measure of response. The former mea-ure gives curves of response vs. velocity that correlate well with curves of contrast sensitivity vs. velocity, wheras the latter does not. 3. Cortical receptive fields were classified according to the criteria of Hubel & Wiesel. Simple cells were found to prefer lower velocities (mean 2-2 deg sec-1) than complex cells( mean 18-8 deg sec-1). The response of simple cells to stimuli moving faster than 20 deg sec-1 is generally poor; complex cells usually discharge briskly to these speeds. 4. Cells classified as hypercomplex by the end-inhibition criterion were further chara-terized as type I or type II, according to the suggestion of Dreher (1972). Type I units are indistinguishable from simple cells in their velocity tuning, and type II units equally clearly resemble complex cells. These results are therefor consistent with Dreher's sbudivision. 5. Teh selectivity of cells for velocity is variable but can be quite marked. The average selectivities of simple and complex cells are not significantly different. There is an inverse correlation between preferred velocity and the sharpness of velocity selectivity for simple cells; no trend is apparent for other cell types. 6. No clear correlation is observed between the velocity preferances of units and their degree of direction selectivity, or receptive field arrangement. Simple cells with 'sustainef' temporal responses to flashed stimuli tend to prefer slower rates of movement than 'transient' ones, and to be less selective for velocity. 7. The results for different cortical cell-types are compared with the velocity tuning of X- and Y-cells in the lateral geniculate nucleus.