What People Actually Need to Know About Everest's Height
The short answer is 8,848.86 meters. That's the official figure agreed on by Nepal and China in December 2020. Before that, everyone was using 8,848 meters from the 1954 Indian survey, and Chinese teams had long argued for 8,844.43 meters based on their own measurements. The disagreement lasted decades. It still comes up occasionally in trivia and arguments on forums like this one. Everest isn't a static object. The mountain sits on the boundary of two tectonic plates that are still pushing into each other. The Indian plate is moving north at about 5 centimeters a year, and that motion deforms the crust below the Himalayas. Everest is technically growing, maybe 4 millimeters a year, but the 2015 Nepal earthquake is thought to have shifted the peak by a few centimeters in an instant. Whether it got taller or shorter from that event is still debated among geodeticists. The measurement itself is tricky. You can't just throw a tape measure at it. Modern surveys use a combination of GNSS receivers, trigonometric observations, and ground-penetrating radar to account for the snow cap versus the actual rock summit. The 2020 measurement used a device called a radiation gauge to estimate snow density, because the snow pack can be anywhere from a meter to several meters deep and shifts every season. That's why the height isn't just a single fixed number like "five miles" or whatever you see on old textbook maps.
I once helped a friend who was compiling elevation data for a GIS project that overlaid satellite imagery onto terrain models. He pulled the 8,848 figure from an open dataset and built a 3D view of the entire range. When he compared it to a higher-resolution LiDAR survey from a research paper, the peak was showing up about 8 meters too low. The issue wasn't the LiDAR. It was that the SRTM digital elevation model they were using had a known bias in the Himalayan zone caused by the way radar penetrates snow and ice. The workaround was to swap in the NASA MERIT DEM dataset, which corrects for that penetration effect, and apply a local adjustment factor derived from the 2020 survey paper. It took about forty minutes to rerun the reprojection. Without that correction, any area-based calculation downstream—slope angles, drainage basins, you name it—was quietly wrong. That's the kind of thing nobody tells you when you're just looking up How Tall The Mount Everest is on a search engine. The number is simple. Working with it at scale is where it gets annoying.
The Technical Details Most People Skip
There are actually two heights worth distinguishing. One is the ellipsoidal height, which is the raw GNSS measurement relative to a mathematical model of the Earth's shape. The other is the orthometric height, which adjusts that number to mean sea level using a geoid model. The 8,848.86 meter figure is an orthometric height. If you're doing anything that requires sub-meter accuracy over a large area, mixing these two up will give you results that look plausible but are systematically off. Another detail that trips people up is the reference datum. Different countries historically used different tidal stations as their zero point. India used Malvan, China used Qionglai. The 2020 agreement effectively harmonized the two, but older datasets still carry the legacy offsets. If you're merging survey data from multiple sources, checking which vertical datum each one references should be the first thing you do, not the fifth. The snow depth variation alone can add or subtract enough to matter for certain applications. Alpine labs have done repeated radar soundings over the summit ridge and found seasonal changes of 0.3 to 0.8 meters in the snow pack. That's not huge, but it's enough to make a single static number feel misleading if you're publishing something that purports to give "the height" without qualifying what exactly is being measured.
Get the Full Details
What to Use Instead If You Need More Precision
If your work requires heights more accurate than a few meters, don't rely on any single published number. Grab the raw GNSS campaign data from the 2020 survey if you can find it through the Survey of Nepal or the Chinese academy of sciences publications. Pair it with a current geoid model like EGM2008 or the newer regionally refined versions that some European groups have put out for the Himalayan arc. It won't be trivial to assemble, but it's the only way to get something defensible if you're publishing technical work. For most everyday purposes, 8,848.86 meters is fine. Just remember that it's a snapshot, not a law of nature, and the next time a good earthquake hits the region, someone will probably be measuring it again.
