The pupil light reflex response (PLR) has been used as an indicator of visual function in neuro-ophthalmology but its usefulness is limited by the use of large light sources and gross measurement techniques. In this study, sophisticated recording equipment (the P SCAN 100 system) and computer-generated stimuli have been used to study certain aspects of the PLR. Luminance masking techniques have been developed which affect the PLR in different ways. In normal observers, local luminance masking eliminates some but not all of the PLR, while light flux changes over a larger area completely eliminate the PLR. Two components of the PLR have been proposed, based on experiments involving normal observers and subjects with cortical damage and optic nerve damage. The first component appears to involve cortical mechanisms as it is absent in subjects with cortical damage. This component has a high contrast gain, but saturates at contrasts above about 30%. The second component may equate with the classical subcortical pupil pathway. It has a lower contrast gain and exhibits extensive spatial summation, and may be involved with the control of steady state pupil size. The pupil has also been shown to respond to spectral changes. This study measures the pupil colour response (PCR) using the P SCAN 100 system and computer-generated stimuli. Luminance masking techniques were again used to ensure that the pupil was responding only to colour change, and not to any increase in light flux when the stimulus was presented. Significant PCRs were measured for a range of colours in normal subjects, but were not present in dichromats when the stimuli were chosen to fall in the same isochromatic zone as the background for each class of dichromat. PCRs were also investigated in subjects with damage to VI and V4. Small but significant responses were measured when large saturated stimuli were used. It is proposed that these responses may involve subcortical mechanisms. A psychophysical colour vision test has been used to investigate the processing of chromatic signals in normal subjects and patients with damage to the early-stage visual pathways. Two stimulus configurations were used - the first involves the detection of vertical bars defined only by chromatic signals (the 'pattern' test), and the second involves the detection of a colour change of a large target already defined by luminance contrast (the 'colour' test). In normal subjects, there is no difference between the chromatic discrimination thresholds found for the two types of test. In this study, two subjects were investigated who performed worse on the pattern test than the colour test. A group of patients who had previously suffered optic neuritis was also investigated to see if the same pattern of results was obtained, but this group showed a great variation in performance. It was proposed that performance at the two tests, indicating the use of chromatic signals, is mediated by two separable neural substrates. These two neural mechanisms can be affected differently but not specifically in optic neuritis. Type I colour-opponent centre-surround cells could be responsible for performance at the pattern test, while Type II colour-opponent spatially co-extensive cells could mediate colour detection when larger targets are used.