Background: Some cortically blind patients have the ability to discriminate visual stimuli presented within their field defect. This residual visual capacity may or may not accompany any acknowledged awareness of the visual event, termed blindsight type I and II respectively. We have previously reported the presence of blindsight in 8 out of 10 cortically blind patients studied. Purpose: To determine the spatio-temporal properties of residual visual capacities in order to characterise the psychophysical channels mediating blindsight performance. Methods: In a 2AFC experiment we have systematically determined the detection of spatially and temporally modulated Gabor patches presented within the blindfield of 10 cortically blind patients. The spatial and temporal frequency ranges investigated were (0.5-7.0) c/deg and (static - 33) Hz respectively. The effect of stimulus size was also determined by systematically varying the standard deviation of the spatial Gaussian (sz) envelope (range 0.5 - 3 degrees). Results: The spatial channels mediating blindsight in all 8 cases were optimally sensitive to low spatial frequencies (0.5- 2 c/deg.) with no significant responses above 4 c/deg. The temporal channel in 4 out of 5 cases examined had a bandpass response characteristic with optimal response between 5-20 Hz. The stimulus size resulting in above threshold (75%) detection was also variable between 6 cases studied (sz between 0.8 - 2.6 deg.). The two negative cases of blindsight performed at chance level under all stimulus conditions tested. Conclusions: The combination of sensitivities to the low spatial and high temporal frequencies together with large spatial summation indicate that mechanisms mediating blindsight are largely driven by magnocellular pathways. The above study demonstrates the importance of stimulus parameters in eliciting blindsight performance. This in part may be responsible for the small number of cases described previously in the literature.