The standard view, that protanopes and deuteranopes see only varieties of yellow and blue, is often taken to be supported by the reports of people with one colour-blind eye and one normal (or near-normal) eye. Judd’s survey of 89 years of unilateral cases might seem authoritative: ‘The color perceptions of both protanopic and deuteranopic observers are confined to two hues, yellow and blue’, corresponding to 575 and 470 nm (DB Judd, J Research National Bureau Standards, 41: 247–271, 1948). A critical re-examination reveals, however, that the cases do not support Judd’s claim: the original publications indicate in most cases substantial (though limited) red-green discrimination with the dichromatic eye under favourable conditions (e.g. with large fields), and evidence of red/green sensation; in the main case where researchers find sensation only of yellow and blue, it is by using experimental methods that are almost guaranteed not to reveal the full range of the subject’s sensation. More recent reports (from the 1950s and later) show that, if the colour experience space of a dichromat eye collapses to a 2D slice through the 3D space of a normal trichromat, the slice may be not flat but curved (and curved in more than one dimension): there are further reasons to think that dichromat colour experience, whether unilateral or bilateral, may not collapse to a plane at all. The unilateral cases are unclear in many ways, but they certainly give no support to a ‘yellow-and-blue’ view of dichromat experience.