Purpose: The ability in cortically blind patients to respond to visual stimuli without consciously experiencing has been termed ‘blindsight’. The goal of this study is to investigate the involvement of the superior colliculi (SC) in mediating blindsight in hemispherectomized (HS) subjects. Methods: We used the achromatic properties of collicular cells, which receive no input from S-cones, to design a reaction time test of a spatial summation effect (SSE), in which reaction times to two bilaterally presented stimuli are significantly faster compared to a single one. Achromatic (SC-visible) and blue/yellow (SC-invisible) gabor patches (1cpd, spatial σ=1cycle, temporal σ=250ms) were displayed on a CRT monitor and isolated either the achromatic or the S-cone opponent (blue/yellow) pathway. Stimuli were presented 10° in the right, left, or in both visual fields. Stimulus onset-time was randomized at 0/500/1000ms with an ITI of 2000ms. Fixation was monitored with an eye tracker. Three HS subjects with and two without blindsight were tested. Results: Subjects with blindsight showed a SSE to achromatic stimuli confirming the presence of blindsight. Presentation of blue/yellow stimuli, however, failed to alter reaction times demonstrating that their blindsight is colour-blind for blue/yellow stimuli. HS subjects without blindsight showed no SSE to either achromatic or blue/yellow stimuli. Conclusions: Blindsight, at least in HS subjects, is color blind, and hence we conclude that it is mediated by the SC. Furthermore, by testing HS subjects we could reject the possibility that spared islands of visual cortex subtend blindsight.