BACKGROUND: In this study, monocular thresholds and binocular summation for abrupt onset/offset versus gradually revealed phantom letter E (illusory contours) stimuli are compared to determine the suitability of these stimuli for assessment of the integrity of two of the major retinal streams: the magnocellular and the parvocellular pathways. Such tests are important in progressive retinal disease where disease severity may differ between the classes of retinal ganglion cells and between the two eyes. Abrupt onset phantom contours have long been considered to activate the magnocellular visual pathway and we propose that gradually revealed high contrast ramped onset/offset stimuli are more likely to promote the more sustained processing of the parvocellular stream. METHODS: Contrast discrimination thresholds for monocular and binocular viewing were compared in a counter-balanced order in 70 young normal subjects, using tests of contrast threshold for a flicker-defined letter E produced by alternation of light and dark dots. Three onset/offset conditions were used - abrupt onset that was maintained for 34 milliseconds (four frames of 8.5 milliseconds) then discontinued, ramped onset over 34 milliseconds (four frames) with offset over 34 milliseconds and ramped onset over 85 milliseconds (10 frames) with offset over 85 milliseconds. RESULTS: Contrast thresholds for identification of the orientation of the E, when presented with four frames ramped onset and offset when compared to the four frames abrupt onset/offset were three times higher, irrespective of monocular or binocular viewing conditions. Threshold contrasts were seven times higher when the 10 frames ramped onset/offset stimuli were compared to abrupt four frames onset/offset. Binocular contrast thresholds were reduced by approximately 40 per cent compared to monocular thresholds for all conditions. The binocular increase in contrast sensitivity is approximately equal for abrupt transiently presented stimuli and for gradually presented more sustained stimuli. DISCUSSION: The results indicate that the same mechanisms of monocular processing and binocular summation are used for identification of a flickering contrast-defined phantom contour under presentation conditions, which are characteristic of the temporal and contrast preferences of the primate magnocellular and parvocellular visual pathways. This suggests that the phantom contour E test may be useful for clinical differentiation of the integrity of the M and P retinal ganglion-derived visual pathways, regardless of whether it is applied monocularly or binocularly.