The McCollough Effect: Why This Image Can Lock a Color Into Your Cortex for Months
Staring at green and red striped grids can trick your brain's primary visual cortex into projecting phantom colors onto black-and-white lines for days, weeks, or even months.

Elena Vasquez-Marsh · for The Unintuitive Universe · October 5, 2026
And it’s been measured. Every claim traced to the published research. Method & sources ↗
Stare at a bright red patch for thirty seconds, then look at a white wall. You will see a faint, greenish glow. This is a standard negative afterimage, a basic trick of the sensory nervous system. It occurs because the photoreceptors in your retina have temporarily fatigued. Blink a few times, wait a minute, and the illusion is gone.
Now consider a different experiment. Instead of a solid patch, you alternate your gaze between two images for ten to fifteen minutes. The first is a pattern of horizontal black-and-red stripes. The second is a pattern of vertical black-and-green stripes. When you finally look at a plain black-and-white grid, the horizontal white stripes appear tinted with green, and the vertical white stripes appear tinted with red.
If you walk away, sleep on it, and check again the next morning, the phantom colors are still there. If you avoid looking at striped patterns to prevent resetting your visual system, you can still perceive these colors on black-and-white grids weeks later. In a controlled study published in the Journal of Experimental Psychology, researchers Allan K. Jones and Dennis H. Holding (1975) demonstrated that just fifteen minutes of exposure can produce a measurable effect up to eighty-five days later.
This is the McCollough effect. Discovered by American psychologist Celeste McCollough in 1965, it is not a retinal afterimage. It is a structural recalibration of the primary visual cortex.
The Cortical Lock
Retinal afterimages are tied to specific spots on your eye. If you tilt your head, the afterimage tilts with you. The McCollough effect behaves differently. If you tilt your head forty-five degrees while looking at the test grid, the illusory colors disappear. If you tilt your head a full ninety degrees, the colors reverse: the stripes that were horizontal on your retina are now vertical, and the pink and green hues swap places.
This dependency on orientation proves the effect does not live in the photoreceptors of the retina. Instead, it occurs in the primary visual cortex (V1), where neurons are specialized to respond to specific orientations of lines. Some V1 cells fire only when they see horizontal edges; others fire only for vertical ones.
The McCollough effect is what vision scientists call a "contingent aftereffect." The brain is not just adapting to a color or a shape in isolation. It is binding them together. The neural population that processes horizontal lines has coupled its activity with the color green to compensate for the heavy red exposure, while the vertical-processing population has coupled with red to compensate for the green.
The physical machinery is highly localized. If you perform the induction phase with your left eye covered, and then view the black-and-white test grid with only your left eye, you will see no illusory colors. The adaptation is monocular, locked into the early cortical pathways where signals from each eye remain segregated.
An Error-Correcting Device
Why would the brain harbor a mechanism that allows a brief visual pattern to distort color perception for months?
The leading explanation, championed by vision researchers like Charles F. Stromeyer III (1978), is that the McCollough effect is a side effect of a critical, permanent calibration system. The human eye is a flawed optical instrument. Because of chromatic aberration, the eye's lens bends different wavelengths of light at slightly different angles. This structural imperfection means that high-contrast black-and-white edges naturally project tiny color fringes onto the retina. Crucially, these fringes are orientation-specific.
If the brain did not compensate for this, the edges of every object in our environment would appear bordered by thin, colored halos. To keep our vision sharp, V1 neurons maintain a slow, constant error-correction process. When the visual system detects a persistent correlation between edge orientation and color, it assumes this is an optical defect of the eye rather than a feature of the outside world. It subtracts the color from those specific edge detectors to restore a neutral, white baseline.
When you stare at Celeste McCollough’s alternating red and green grids, you feed this calibration system highly artificial data. The brain concludes that your eye’s lens has suddenly developed a massive chromatic aberration. It deploys its permanent correction program, subtracting red from horizontal edges and green from vertical ones.
Unlearning the Illusion
Because this error-correction mechanism is designed to fix permanent physical properties of the eye, it does not decay passively with the simple passage of time. A study by Edward Vul, Janet Krizay, and Donald I. A. MacLeod (2008) confirmed that the McCollough effect consists of two distinct components: a rapid, transient adaptation that fades over seconds, and an essentially static, long-term bias that does not decay in the dark.
If you go to sleep in a completely dark room, your visual cortex holds onto the calibration settings. The only way to truly erase the effect is to actively expose the system to new, conflicting environmental statistics. Every time you look at normal, uncolored black-and-white lines in your daily life, the visual system receives data showing that horizontal and vertical edges are, in fact, colorless. This input gradually counter-conditions the effect, forcing the neural networks in V1 to update their weights once again.
Until that unlearning process is complete, the neural adjustment remains. The phantom pinks and greens on the grid are a physical record of your brain attempting to heal an optical defect that never existed.
Measured.
This article is AI-generated (synthetic) content, produced by an automated editorial system with human direction and review. Every claim is traced to published, peer-reviewed sources.