The Face on Mars
DebunkedSummary
On 25 July 1976, the Viking 1 orbiter photographed a hill in the Cydonia region of Mars, and in that one photograph, at that one sun angle, at that one resolution, it looks like a human face staring up out of the desert.
It is a hill.
The most satisfying detail in the case is this. In the Viking image, the face appears to have a nostril.
There is no nostril. It is a missing pixel. The data packet was lost in transmission, and the gap in the image happened to fall in exactly the right place.
What is documented
The image. Viking 1 orbiter, frame 35A72, taken on 25 July 1976 over Cydonia Mensae.
The resolution. Roughly forty-three metres per pixel. Each pixel in that image covers an area about the size of a tennis court.
The lighting. The sun was low. Shadows were long, and they were long across a landform with an irregular top.
The data dropouts. The image, like many Viking frames, contains black speckles: pixels where the data was lost in transmission from Mars. One of them sits where a nostril would be.
NASA released it. The agency published the image with a caption noting the resemblance and attributing it to shadow. It was not hidden and it was never hidden.
What was built on it. A claim that the landform is an artificial monument, accompanied by claims about neighbouring features described as a city and a pyramid.
What happened when we looked again.
April 1998. Mars Global Surveyor photographed Cydonia at roughly four metres per pixel: ten times the resolution of Viking. It is a mesa.
April 2001. Mars Global Surveyor again, with better lighting and better geometry. It is a mesa.
2007. The High Resolution Stereo Camera on ESA's Mars Express imaged it in three dimensions. It is a mesa, and now we have its shape.
2010. HiRISE, on the Mars Reconnaissance Orbiter, imaged it at around thirty centimetres per pixel: more than a hundred times the resolution of the original.
It is a mesa. It is a perfectly ordinary eroded landform of a kind that is abundant in Cydonia and abundant on Earth.
Leading explanations
A mesa, plus pareidolia, plus a low sun, plus a missing pixel. Settled beyond any dispute.
On pareidolia, and this is the part worth taking seriously.
The human brain has dedicated hardware for recognising faces. There is a region, the fusiform face area, whose job is to detect them, and it is astonishingly sensitive, and it is tuned to fire on the barest hint of two dots and a line.
That sensitivity is not a defect. It is one of the most valuable pieces of processing we possess, and it evolved because failing to notice a face was, for most of our history, a great deal more expensive than noticing one that was not there.
The consequence is that we see faces in electrical sockets, in car radiators, in clouds, in burnt toast, in the grain of wooden doors, and in a hill on Mars photographed at forty-three metres per pixel with the sun low behind it.
Seeing the face is not a mistake. It is the system working exactly as designed. Believing the face is there is the mistake.
What the popular version gets wrong
"NASA covered it up." NASA released the photograph, in 1976, with a caption drawing attention to the resemblance, and then spent thirty years photographing it repeatedly at higher and higher resolution and publishing every image.
"The newer images were doctored to hide it." Four separate imaging systems, on three separate spacecraft, from two separate space agencies, over thirty-four years, in visible light and in stereo, at resolutions from four metres down to thirty centimetres.
"It has a nostril, which erosion cannot explain." It has a lost data packet.
"At low resolution it looks artificial, so something is being hidden at high resolution." At low resolution, with a low sun, a great many things look artificial. That is what low resolution does. The correct response is to take a better picture, and we took a better picture, four times.
Current status
Debunked, and about as comprehensively as anything can be.
The registry keeps this record for what it teaches about looking, which is a subject this whole archive is about.
The Face on Mars was never a question about Mars. It was a question about a photograph, and it was settled the moment somebody took a better one. What kept it alive for twenty-two years was not the absence of an answer. It was that the first picture was so much more interesting than the second.
Sources
- NASA, Viking 1 orbiter, frame 35A72, 25 July 1976, and the accompanying press release.
- Mars Global Surveyor, Mars Orbiter Camera images of Cydonia, April 1998 and April 2001.
- ESA Mars Express, High Resolution Stereo Camera imaging of Cydonia, 2007.
- NASA Mars Reconnaissance Orbiter, HiRISE imaging of Cydonia, 2010.
- Literature on pareidolia and on the function of the fusiform face area.
Last reviewed: July 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.