How Do Dogs Actually See Color? The Truth About Dog Vision

Published August 2026 · Pet N Pet Editorial · National Dog Day Science Series

Start by throwing out the thing everyone repeats: dogs do not see the world in black and white. That myth has been dead in the research for decades. Dogs see color. They just see a narrower, two-channel version of it. Back in 1989, scientists established that the dog retina has two types of color-sensing cone cells, not the three humans have, making dogs what's called dichromats (Neitz, Geist, and Jacobs, Visual Neuroscience, 1989).

In practice, that means a dog's world is painted mostly in blues, yellows, and grays, with reds and greens washing out and blurring together. It's roughly the same palette a person with red-green color blindness sees. Here's what science has actually measured, and why it means the red ball you throw onto green grass may as well be invisible.

Two Cones, Not Three: What That Actually Means

Human color vision runs on three types of cone cells, tuned to roughly red, green, and blue light. Mixing their signals is what lets us see the full rainbow. Dogs have only two cone types, with peak sensitivities around 429 nanometers (in the blue-violet range) and 555 nanometers (yellow-green), a finding confirmed by both behavioral testing and later measurements of the photopigments themselves (Neitz, Geist, and Jacobs, 1989; Jacobs et al., Visual Neuroscience, 1993).

With two cones instead of three, a dog's color world collapses from a full spectrum into essentially two hues plus everything in between: blue and yellow. There's also a specific wavelength, around 480 nanometers, that looks plain gray to a dog, with no color at all (attributed to Neitz et al., 1989). So dogs aren't seeing a dull gray world. They're seeing a real but limited color world, one missing the entire red-green dimension that we take for granted.

Which Colors Can Dogs See, and Which Ones Vanish?

Blues and yellows come through clearly for dogs. Those sit right where their two cones are most sensitive. The trouble is everything in the red-to-green band. To a dog, a bright red object doesn't glow red, it reads as a dark, muddy brownish-yellow or even grayish. Green tends to wash toward that same yellowish middle. So red and green, which look boldly different to us, look confusingly similar to a dog.

A 2017 study drove this home with a dog version of the Ishihara color-blindness test used on people. The dogs behaved just like red-green colorblind humans, unable to tell certain red and green shades apart (Siniscalchi et al., Royal Society Open Science, 2017). This is the practical headline for owners: your dog can see blue and yellow well, but it struggles badly with reds and greens.

Do Dogs Even Care About Color? More Than We Thought

For a long time the assumption was that dogs mostly ignore color and rely on brightness, how light or dark something is. A 2013 experiment tested that directly. Eight untrained dogs were given cues that differed in both brightness and color, so they could have used either. Given the free choice, the dogs went with color (Kasparson, Badridze, and Maximov, Proceedings of the Royal Society B, 2013).

The authors concluded that in normal daylight, color information may actually be the dominant cue for dogs, even with just two cone types. That overturned the old brightness-first assumption. It also has a real-world payoff: when you're picking a toy your dog needs to find, the color genuinely matters to the dog, not just to you.

Sharpness, Darkness, and Motion: The Other Trade-Offs

Color is only part of how a dog sees. In fine detail, dogs come up short. Their visual acuity is commonly estimated around 20/75, meaning what a dog can just make out at 20 feet, a typical person could see clearly from about 75 (Miller and Murphy, JAVMA, 1995). That figure is an average, though, and estimates across methods range widely, so treat it as a ballpark rather than a fixed number.

The trade-offs run the other way in dim light and motion. Dogs have rod-rich retinas, the cells built for low-light vision, which is why they navigate the yard at dusk far better than you do. They're also highly tuned to movement. In evolutionary terms it's a sensible package for a crepuscular hunter: give up some fine detail and color range, and gain the ability to detect a moving animal in near-darkness

So What About Those 'This Is What Your Dog Sees' Images?

You've probably seen side-by-side photos claiming to show a scene through a dog's eyes, all muted blues and yellows. They're useful, but be clear on what they are. Those images are built by mathematically remapping a normal photo into a two-cone color range and blurring it to approximate dog acuity. They're educated models based on the cone data, not recordings of what a dog actually experiences.

No one has ever measured a dog's private, subjective color experience, and no one can. The simulations are our best reconstruction of the dog's likely color gamut, which is a genuinely useful thing. Just don't mistake an estimate for a photograph taken from behind your dog's eyes. The honest phrasing is "this approximates what a dog probably sees," not "this is what your dog sees."

Frequently Asked Questions

Can dogs see color, or only black and white?

Dogs see color. The black-and-white idea is a myth. Their retinas have two cone types instead of the three humans have, making them dichromats (Neitz et al., 1989). That gives them a real but limited color world built mostly around blue and yellow, with reds and greens washing out, similar to a person with red-green color blindness.

What colors can dogs not see well?

Reds and greens. With only two cone types, dogs can't distinguish the red-green dimension, so a red object tends to read as dark brownish-yellow and green washes toward the same muddy middle. A dog version of the Ishihara color test found dogs behaved like red-green colorblind humans (Siniscalchi et al., 2017).

What color toys are best for dogs?

Blue and yellow. Those sit right where a dog's two cones are most sensitive, so they stand out clearly, while red and green blend into the background. That matters more than owners assume, because dogs actually rely on color to make choices (Kasparson et al., 2013). A red ball on green grass is one of the hardest things for a dog to spot.

Is a dog's color vision like human color blindness?

Functionally, yes. A dog's two-cone vision is close to human red-green (deuteranope-type) color blindness. Both see blues and yellows fine and confuse reds and greens. It's not a defect in dogs, though, it's simply how canine vision is built, and it comes packaged with better dim-light and motion vision than humans have.

Do dogs see better than people in the dark?

Yes. Dogs have rod-rich retinas optimized for low light and are highly sensitive to motion, which is why they handle dusk and movement far better than we do. The trade-off is fine detail and color range: dog acuity is estimated around 20/75 on average (Miller & Murphy, 1995), less sharp than typical human eyesight.

Bottom Line

Dogs live in color, just a simpler, two-note version of it. Blues and yellows come through clearly, reds and greens blur into a muddy sameness, and a specific band of blue-green looks like plain gray. It's essentially the world as seen through red-green color blindness, paired with sharper night and motion vision than we'll ever have. The practical lesson is small but real: buy the blue toy, not the red one, and don't be surprised when your dog trots right past a crimson ball sitting in the grass. To its eyes, that ball was never there.

SOURCES & CITATIONS

Neitz, Geist, Jacobs  "Color vision in the dog," Visual Neuroscience 3(2):119-125 (1989). Established dichromacy; cone peaks ~429 and ~555 nm; neutral point ~480 nm.  https://www.cambridge.org/core/journals/visual-neuroscience/article/abs/color-vision-in-the-dog/5A30E35A384D10B6A2B46B01C896D3D5

Jacobs, Deegan, Crognale, Fenwick  "Photopigments of dogs and foxes and their implications for canid vision," Visual Neuroscience 10(1):173-180 (1993). Cone peaks ~430-435 and ~555 nm; rod peak ~508 nm.  https://pubmed.ncbi.nlm.nih.gov/8424924/

Kasparson, Badridze, Maximov  "Colour cues proved to be more informative for dogs than brightness," Proceedings of the Royal Society B 280(1766):20131356 (2013). 8 dogs chose by hue over brightness.  https://pmc.ncbi.nlm.nih.gov/articles/PMC3730601/

Miller, Murphy  "Vision in dogs," JAVMA 207(12):1623-1634 (1995). Canonical veterinary review; visual acuity estimated ~20/75; rod-rich, motion-sensitive retina.  https://avmajournals.avma.org/view/journals/javma/207/12/javma.1995.207.12.1623.xml

Byosiere, Chouinard, Howell, Bennett  "What do dogs (Canis familiaris) see? A review of vision in dogs...," Psychonomic Bulletin & Review 25(5):1798-1813 (2018). Corroborates the ~20/75 acuity estimate (range ~20/50-20/140).  https://link.springer.com/article/10.3758/s13423-017-1404-7

Siniscalchi, d'Ingeo, Fornelli, Quaranta  "Are dogs red-green colour blind?" Royal Society Open Science 4(11):170869 (2017). Dog Ishihara-style test; dogs behaved like red-green colorblind humans.  https://royalsocietypublishing.org/doi/10.1098/rsos.170869