Mt Roraima – Guyana Highpoint

Mt Roraima – Guyana Highpoint

June 26 – July 2, 2026

50 miles, 10kft gain

Eric Gilbertson and Andres Rodriguez

Survey results:

I determined the true highpoint of Guyana is at location (5.210989, -60.739325) with elevation 2805.8m +/-0.1m EGM2008. This is also the true summit of Mt Roraima which, contrary to common belief, is in Guyana.

Maverick Rock elevation: 2754.20m +/-0.06m

June 25 – Fly to Boa Vista, taxi to Pacaraima
June 26 – cross border to Venezuela, drive to Paraitepuy
June 27 – hike to campo base
June 28 – hike to principle hotel camp, survey Maverick Rock
June 29 – hike to triple country point, camp at Coati Camp
June 30 – climb and survey Mt Roraima summit, Guyana highpoint
July 1 – hike to principle hotel camp
July 2 – hike to Paraitepuy
July 3 – drive to Santa Elena

Guyana was one of the few remaining countries in the world where the exact location of the country highpoint was not known. Border location uncertainty and challenging terrain make determining the exact location difficult.

Existing ground surveys from the early 1900s put the highest point somewhere on the top of the Mt Roraima tepui near the Venezuela border, but the border is disputed. Knowing the correct border location is important in determining the correct location of the Guyana highpoint.

It would be easy to assume that current online map resources like Google maps have country borders all drawn correctly, but I’ve discovered this is not the case. For instance, I previously discovered Google maps had the Chad-Libya border incorrect by 500m when I compared it to the actual treaty definition. I contacted the team lead at Google for that area and showed them the error and they agreed with my analysis and fixed it. Needless to say, now I do my own homework to find borders if it is important for a country highpoint.

Guyana was formerly a British territory, and an 1899 international court ruling called the Paris Arbitral Award established the border between Venezuela and British Guiana. In 1966, Guyana became independent and this became the international border.

Venezuela does not agree with the court ruling and claims the Essequibo region west of the Essequibo River as Venezuelan territory. This is still being litigated at the United Nations. However, until a ruling is made, the internationally-recognized border is still that of the 1899 ruling.

The 1899 ruling states that the border in the area of Mt Roraima is defined as “…following the mid-stream of the Wenamu [Venamo] to its westernmost source and thence in a direct line to the summit of Mt. Roraima…” Knowing the exact location of this border is critical in determining the highpoint location.

In 1905 a ground survey was conducted to place boundary markers at the two endpoints of this straight line border. According to the field notes, on Jan 10 a monument was placed at (5 56 55.4, -61 23 24.7) to mark the western source of the Venamo River, and another monument at (5 10 09.6, -60 45 58.2) on Mt Roraima. Note that for boundaries, monuments on the ground supersede written boundary definitions, so even if the boundary monument on Mt Roraima was not on the true summit, that monument is the actual border.

A 1931 survey on Dec 29 gave updated coordinates for the Roraima monument as (5 12 08.6, -60 44 07.5) from a Venezuelan commission and (5 12 18.92, -60 44 02.31) from a Brazilian commission. Some of the differences may be due to errors in surveys, and the most recent survey supersedes the older surveys.

The coordinates are all in the La Canoa datum, so cannot directly be entered onto an online map that requires WGS84 datum. I used an online conversion tool from epsg.io (https://epsg.io/transform#s_srs=4247&t_srs=4326&ops=1516&x=NaN&y=NaN) to convert La Canoa to WGS84 and found the Western monument is at (5.9455472, -61.3918846). 

For the Roraima monument, the three different coordinates are all at slightly different locations, so it’s unclear which is correct. However, given that the physical monument on the ground is the ultimate definition of the boundary, I figured maybe I could find someone that had gone there with a GPS more recently to get accurate coordinates.

Luckily the Roraima triple country boundary monument is a popular hiking destination, and I found Sean Caulfield on peakbagger had logged coordinates of the monument in 2016 as (5.2019,-60.7375). Even though this was a handheld GPS that is not as accurate as a survey grade GPS, these would likely be the most accurate coordinates I could get and would likely be within 10m of the correct location. I tried to find reports of anyone visiting the next monument at the western source of the Venamo River, but was not successful. So I would just use the converted coordinates from the 1905 survey.

I next plotted these points in caltopo and connected them with a line. Not too surprisingly, all online maps were slightly in error. They either have the coordinates slightly off for one or both of the two monuments or, in some cases, add a kink in the border at the edge of the Roraima tepui. There should obviously not be a kink since the 1899 treaty states it’s a straight line border.

I also double checked with LSIB (Large Scale International Boundaries) (https://geodata.state.gov/geonetwork/srv/eng/catalog.search#/metadata/3bdb81a0-c1b9-439), the US government source for international boundaries. This is what I’ve been told Google maps uses, but they aren’t always consistent. I downloaded the Guyana boundary and opened in QGIS, specialized surveying software. This border was also slightly off. Though the documentation for LSIB advises up to +/-100m errors are possible. I found a 160m error at the Roraima monument, which is about what could be expected.

Given all this variation in boundary sources, I placed the highest trust in my own analysis, since it was based on the original treaty, original ground surveys, and recent ground measurements.

With my border drawn on caltopo I started looking at where different Digital Elevation Models (DEMs) put the highest point in Guyana. There has never been a ground survey of this area of Roraima, so the best existing measurements are satellite based. Satellites have taken elevation measurements at 30m horizontal spacing (in the years 2000 and 2021) and DEMs are used to approximate the elevation between sampled points. These are subject to high error if sharp peaks or valleys are missed, but these are the only elevation measurements available for this area.

I looked at the major DEM models like COPERNICUS, ALOS, NASADEM, and a few others. They all put the highpoint directly on the border on the side slope of a peak between the Roraima monument and the edge of the tepui. This was interesting since it would be one of only two country highpoints in the world that are not on exact summits (the other is Monaco).

In theory, if I could get up Roraima I could just walk along the border and use a surveying auto level or sight level and find the highest point. However, it wouldn’t actually be that easy. I learned that the northern area of Roraima is called “The Labyrinth” and is very complicated terrain with deep chasms and towers. This matched what I saw from satellite images. So I likely couldn’t easily walk that border.

My plan was then to try to get to the DEM highpoint candidates with a surveying level to identify which point was highest, or if there was a higher point nearby. Then I would measure that with my GNSS equipment.

Interestingly, it makes the most sense to climb Roraima from the Venezuela side. Roraima is a tepui, meaning it is a flat-topped peak with cliffs on all sides. The one weakness in the cliffs is a ramp on the Venezuela side. This side has a trail up and is a common trek for backpackers. Climbing Roraima from the Guyana or Brazil sides would require a jungle bushwhack approach and long technical rock climb and is not very practical.

Unfortunately, Venezuela had been off limits to Americans since about 2016, so I had to put my plans on the shelf until the political situation improved.

In 2024 local indigenous guide Andres Rodriguez and British mountaineer Charlie Bicknell made the first ascent of a candidate highpoint location in this area I’d identified. Charlie wrote that it visually looked like the highest point around, and he said he took measurements with his GPS watch and a drone to say it was the highest point. No values or error bounds were given for drone measurements.

I’ve learned through years of surveying with professional equipment that GPS watches can have vertical errors up to 50m or more, so they can’t really be used to measure accurate elevations.  Consumer drones rely on similar technology for elevation measurements and are also not accurate enough in cases like these. A lidar drone paired with a GNSS base station would be very accurate, but that wasn’t used.

The candidate location was not consistent with any existing digital elevation models, which raised questions about it being the correct location. So based on the measurements Charlie took it was not determined for sure if he was at the highest point.

In the case of determining a country highpoint I think it is critical to take accurate measurements before making a claim that the peak is the true highest. Appropriate surveying equipment with known error bounds should be used. In this case, the appropriate equipment would at the minimum be a surveying sight level or abney level, and better yet a standard auto level on a tripod. These tools can each accurately determine if rival peaks are higher or lower.

To measure absolute elevation, a survey grade GNSS receiver should be used. These can get 0.1m or better vertical accuracy with known error bounds, unlike a GPS watch which has very high but unknown error bounds. Measurements should be conducted with sufficient rigor such that they could be written up in a scientific journal article and accepted in a geography journal. This is what I’ve done for all the eight new country highpoints I’ve discovered so far (published in Progress in Physical Geography: Earth and Environment Journal, and under review in several other journals).

Based on my experience measuring peaks around the world, I concluded the true highpoint of Guyana was not yet known, and an accurate survey was needed to determine its location and elevation.

In April, 2026, Venezuela began issuing tourist visas to Americans again, so I started making plans for Roraima. If I could get a visa, my plan was to bring a 32x zoom surveying auto level with 20 arcsecond resolution, tripod, and two surveying sight levels to the DEM candidate highpoints and to the candidate climbed by Andres and Charlie. On the summits I would use these tools to measure if surrounding peaks were shorter or taller.

I would then bring two survey grade GNSS units (one primary, one spare) to measure the elevation of the highest identified peak. I would compare the measured lat long coordinates to the known border to verify if it was indeed in Guyana.

My plan was to go during the summer during my  break from teaching. This is the wet season so it is not ideal for climbing or surveying. However, I wanted to get to Venezuela as soon as possible while the window for entry was open, since the political situation could easily swing back and shut the window again for me.

Since Andres and Charlie’s climb there had been three additional teams climbing the candidate highpoint. Alex Hubin, Viri and Fernando, and Andreas Aschaber. They all went with Andres as guide. It is legally required to hire a local guide to hike up Mt Roraima, and the route up the candidate highpoint was reportedly complicated and technical, so I also decided to go with Andres to increase chance of success.

On May 6 I sorted out my summer schedule and sent in my visa application for Venezuela. It was an electronic visa and I’d read on Venezuela Visa Facebook forums that people were getting approved within 10 days. So it seemed like plenty of lead time for my scheduled entry to Venezuela on June 25. Jared would join, and he sent his application in then too.

The visa form was straightforward, just requiring flight confirmations, hotel reservations, and standard questions. I planned to enter Venezuela by land from Brazil by flying to Boa Vista and taking a taxi to the border. This is the standard way to access Roraima since there are no nearby major airports on the Venezuela side.

For the visa form a round trip flight itinerary is required, though. So I just bought a random flight in and out of Caracas on Expedia, saved the receipt, then cancelled for full refund since it was within 24 hours.

One part of the visa was listed as optional to include a sworn signed declaration of no criminal history or an official criminal background check. I just wrote up a few sentences in a word doc, signed it, and submitted that.

On May 13 the visa got rejected and I realized it was not going to be as easy as anticipated. It appears the visa requirements were evolving over time. The people that applied in April didn’t need the sworn statement. But my rejection letter said I now needed an official criminal background check appostilled, notarized and translated. This was a new requirement I’d never encountered before in any past visa applications.

There’s an FBI background check government website I filled out and it requires showing up to a post office in person for finger printing. I tried three separate post offices and none of the finger print machines worked for me. I think it’s because I’d gotten frostbite on all my fingers on Peak Pobeda a few years ago, and that the machines are low quality.

I next decided to pay a private company for the background check. It’s likely they would have a higher chance of success vs a free government service.

I found Certifix would do it quickly. I paid online and they had me go to a UPS store for fingerprinting. I worked with an employee for an hour and we finally got all the fingerprints to work. I submitted them online and paid for appostilling, notarizing and translating services, which would take two weeks.

By June 4 I got the documents back and immediately uploaded them to the visa site. By then a $60 fee had been added to the requirements, but it was unclear if I needed to pay that since it wasn’t required when I first submitted.

I joined a very helpful Whatsapp group of other people trying to get visas. It appeared rejections were now pretty much random, sometimes with no reason given at all, wait times were random, and the whole system was in chaos.

By June 21 I still hadn’t gotten my visa. I found someone in Caracas and paid them to go in person to the office and check. They said my visa was in a queue and I’d just have to wait.

On June 22 I flew out to the first country on my trip, Jamaica, even though I still didn’t have the Venezuela visa. On June 23, while I was surveying three highpoint candidates in Jamaica, I started making backup plans with Jared since it was only two days before we planned to enter Venezuela and we still hadn’t gotten our visas. I had cell service on the summit of Blue Mountain Peak, and as I was texting with Jared I got an email with my visa approved and attached!

Jared still hadn’t gotten his, but we decided I would continue as planned since the Roraima logistics and guide were already arranged and paid for.

I flew out the next day to Manaus, and that evening, June 24, a big earthquake hit Caracas. I had a night layover on Manaus, and I reused my existing Brazil visa from December to get in the country. I slept in the airport a few hours, then flew to Boa Vista the next afternoon.

In Boa Vista I found a taxi to take me to the border at Pacaraima. I had been optimistic I might be able to cross before it closed at 6pm, but a flight delay and a delay waiting for the share taxi to fill up meant I didn’t get to Pacaraima until 645pm. So I stayed the night at the Pacaraima hotel.

The next morning, June 25, I got in line for the leaving-Brazil immigration office at 630am. They opened at 8am and were very efficient and soon had stamped my passport.

I then walked back to Pacaraima and got in a share taxi to get to the Venezuela side, since the officers wouldn’t let me walk across.

Three locals got in the taxi and we drove across. Nobody on the Brazil side checked any documents. Then we got to the checkpoint to enter Venezuela. We slowed down to a stop at a few armed officers. The officers then waived us through and we kept driving. We passed one more checkpoint and the military officers just waived us through.

Amazingly, I had made it in to Venezuela without having my passport or visa checked at all! It appears locals move back and forth for work between Santa Elena and Pacaraima and they don’t ever need to go through immigration checks. Since I was in a share taxi full of locals, the officer didn’t notice me and just waived the whole car through. The taxi driver knew I was not local, but he didn’t want to stop and have to wait for me to get stamps. So he just drove through.

I met Antonio, my guide, in Santa Elena and told him what happened. He said it would cause problems for me in the future at checkpoints and airports in Venezuela if I didn’t have the stamp and in my passport. So I left my luggage with him and took a taxi back to the border, about 10 minutes away.

I explained the situation to the Venezuela immigration officer and he smiled like that had happened before. He asked me a lot of questions about my plans and had me state my plans while he video recorded me with his phone. I’ve never had to do that before at a border. I think I’m the first American to pass through that border for a long time, so things were a little different for me.

He wanted to inspect my luggage, but since it was all in Santa Elena that wasn’t possible. This perhaps worked out well for me since I had a bunch of surveying equipment that might raise eyebrows, especially since I was planning to visit the contested border zone. Luckily that topic never came up.

The agent then made some calls. He then told me,  because of the earthquake, he couldn’t log in to his system, and he couldn’t get his supervisor from Caracas on the phone. He said I could not enter Venezuela and would have to come back and try again next week.

To be certain I understood correctly with my limited Spanish, I asked if he could type what he said in the translation app on my phone. Then he typed in that I should try coming back in December!

I told him my luggage was all in Santa Elena, and he said I could taxi in to get it but had to come straight back.

When I walked outside, Antonio had driven up there and I explained the situation. So he went in to talk to the agent himself, and explained that the trip was already organized and paid for and ready to go, and that I would come straight back to the border after the trip to get the stamp and visa.

The officer agreed this was ok, and let me go back into Venezuela, even though my passport was not stamped and I wasn’t entered in their online system.

We loaded up the Land Cruiser with all my gear, and Antonio, Omar, and Marcelo joined. Antonio was the lead guide, Omar was the cook, and Marcelo was a porter. We drove an hour on paved roads to San Francisco village, ate lunch at a restaurant, then drove another hour on a dirt 4×4 road to Paraitepuy village. There we checked in with the park staff and set up tents for the evening. Roraima was visible in the distance.

June 27

The next morning we started our trek, hiking for three hours on a good trail through open grassland to Rio Tek, the last outpost of civilization. A few families live there and sell food to trekkers. The outpost can be reached by motorcycle for $40 each way and it takes about 30 minutes. Andres had a motorcycle and met us there.

We stopped for lunch, then continued hiking. We passed a small church across the river up a hill, then crossed one more river and started ascending. It started raining later and we reached campo base in the rain just before sunset.

This camp is at the edge of the jungle at the base of the cliffs of Roraima. It appears the area below there has historically been cleared by fire so the jungle got replaced by grassland for cattle grazing.

June 28

We took our time the next morning hiking up the ramp that allows passage to the top of the tepui. The tepui is composed of huge vertical cliffs all around, that are sometimes overhanging. It would be very challenging to get up except for the one ramp that amazingly is big enough for a hiking trail.

We hiked through jungle and through a few waterfalls, and topped out on the tepui after four hours.

The top looks like a different world. There are carnivorous plants and swampy areas, but very few trees. I think it is generally very windy up there and a harsh environment. There are a few rock towers near the entrance and there are small trees and shrubs in the sheltered areas of these towers.

The only animals appear to be birds, mice, and coati, a small carnivorous mammal the size of a Chihuahua.

We were the only group up there so got to claim the biggest camp site, the “Principle Hotel”. The rock up there is all sandstone which gets eroded away to form overhangs on the edges of the towers. These provide shelter from the rain and are the main campsites. A half dozen established “hotels” are designated, and groups are required to camp at designated places.

Our place had room for all four tents, and we dropped gear and had lunch. I then hiked up nearby Maverick Rock with my survey equipment.

Maverick Rock is interesting because it has historically been considered the highest point of the Roraima tepui. It is also conveniently right next the the access ramp. Charlie said his GPS watch read the candidate highpoint in Guyana was taller than Maverick Rock. But GPS watches are subject to such high errors that they can’t be trusted for any consequential measurements within 50m or so.

It’s a bit nebulous exactly how GPS watches spit out an elevation, but likely a built in barometer plays a role. With weather systems passing through Roraima so frequently it is easy for pressure changes to add huge errors to any measurements. The watches also don’t report which vertical datum is being used, which adds to uncertainty.

Since I was there I decided to take an accurate elevation measurement. I brought up the 32x auto level and set it up on a tripod. Interestingly, I couldn’t bring my regular surveying tripod on my trip since it would have counted as oversized luggage and made flying more complicated. So I had purchased a smaller tripod that would accept an antenna rod.

That would be useful for my Jamaica surveys in the jungle that required a long antenna rod, plus the tripod could also work for the auto level.

I mounted the level on the top of Maverick Rock and verified all other surrounding points on that side of the tepui were indeed shorter. So it was at least the tallest point on the Venezuela side. The DEMs were all in error about that, since they showed other points to the NE and SW much taller. The DEMs on Roraima in fact have the largest errors I’ve ever seen in the world.

I then took a one hour measurement with a Trimble DA2. I would later post process the result to find the elevation 2754.20m +/- 0.06m EGM2008. This was about 30m taller than my Garmin Fenix 6 GPS watch was reading, and the watch fluctuated by +/-10m during the reading. This goes to show how unreliable a GPS watch is for measuring elevation.

Clouds rolled in and out, and I never got a clear view of the Guyana side of the tepui. So that would have to wait.

June 29

The next morning we hiked about 3 hours deeper onto the tepui, reaching the triple country boundary marker around 10:30am. One side has Brazil written, one side Venezuela, and the third side is blank. I’ve heard a venezuelan colonel ripped off a plaque that said Guyana, likely because Venezuela contests that border.

On the northwest corner of the structure is a small brass monument. I mounted the DA2 there and took data for half an hour. I would use this to correct my previous coordinates for the border. It started raining halfway through but I covered the battery in a bag and got the data. The coordinates of the monument are (5.201889, -60.73755).

We then hiked another 45 minutes through the rain and reached Coati camp. This is located in an interesting little cave that leads to an area open to the sky above with trees growing. The cave is small but very sheltered from the wind and rain. We hung out there until the rain abated, then went outside and got a good view of the highpoint candidate nearby.

It looked like complicated terrain to access, and a nearby peak with a balanced rock on top looked about equally tall. The DEM points were behind it out of view, so it was hard to tell what was actually the highest point.

Andres said the previous climbs had all taken about 1.5 hours from camp, but they were all in the dry season. Things would be more challenging now in the rainy season.

June 30

The next morning started socked in with fog as usual, with light mist ensuring all rocks were soaked. It wasn’t actively raining, though. I’ve found weather forecasts from my inreach were unreliable up there and rain can pretty much happen randomly any time of day for random amounts of time. There’s no pattern of clear in the morning and rainy in afternoon like on many other mountains.

So we started up at 730am hoping for the best. Right out of camp we climbed a short 15ft vertical step that was low 5th class. Andres soloed up and belayed me following. I didn’t want to take any chances while hauling up a big pack full of survey equipment climbing slimy wet rocks.

Above the step we hiked across flat slabs, then bushwhacked through a short shrubby section to the base of the candidate tower. From there the route got complicated and I can see why the area is called the Labyrinth. We squeezed through narrow slot canyons, jumped boulders across deep chasms, scrambled up wet slimy narrow rock fins, and pulled our way up dirt slopes using small roots. Even with a GPS track to follow it would be tricky to repeat that route.

We eventually got to the base of a small arch, which required a delicate slab traverse with a 10ft drop below. I needed to make a long reach with my right arm to grab a jug, then pull hard on my right arm. That’s my bad arm, where I’d dislocated my shoulder in the past on a fall on Mt Everest. When I made the move I could feel my shoulder pop, but luckily it stayed in place.

We then passed under an arch into a tall narrow slot canyon. I took off my pack there to push ahead of me as I squeezed through. When the canyon ended at a steep dropoff there was a big chimney to the left. Andres got the rope out there, and trailed it while he climbed up.

At the top he slung a horn and gave me a belay. The chimney was tricky since it was soaking wet and slimy on the sides, and I had a big pack full of equipment. Halfway up I had Andres drop a rope end down and he hauled the pack up on its own. Then I wriggled up to just below him, and tugged on a rope end to get past the final slimey lip.

We scrambled a bit unroped from there, then reached a slimey slabby face. Andres soloed up again and I don’t see how he didn’t slip on the slime. The face was steep. He had rubber boots and I had hiking boots, so his boots appeared more appropriate and grippy perhaps. Also he didn’t have a big pack on. But he’s probably just a stronger climber than me.

As I started up the face on top rope it started raining hard. I tried to reach right to grab a good hold, but slipped on the slime and fell. My shoulder dislocated in the fall but the rope held me. I quickly got my shoulder popped back in while dangling on the rope, then decided there was no glory to be had in trying to free climb the slimey face. It was best to just get up and out of the rain.

Andres threw me a rope end and I pulled myself up. We then hustled over to a narrow overhang where we could barely huddle out of the rain standing up against the cliff.

We stood there shivering for the next hour. One thing about the rain up on Roraima is that it generally will have gaps between downpours instead of being continuous all day. So we just had to wait.

After an hour the rain let up. We retrieved the rope, then Andres soloed up the next 20ft slimey slab and belayed me up. This time I didn’t slip and didn’t need extra rope for aid.

The rain came back again, but at this point we just had one pitch left and no shelter, so we continued. Andres soloed across a traverse slab with the rope anchored on one end. He then slung a horn on top and tied the rope taut. I then clipped on and traversed, then pulled myself up over the last lip.

The rain then let up, and we were past the technical section. We hiked over and reached the highpoint, which was two rock fins of nearly equal height separated by a 15ft deep gap. I tagged both fins, and with my sight level measured the N one was the true summit. This was also the easiest one to reach (note: not the one in summit pictures from the previous trips).

Just then the clouds miraculously cleared, and I had views all around! This was incredible news for the survey, and very unexpected given the morning rain. I quickly looked around and identified at least five other peaks nearby that were too close to call by eye if they were shorter or taller.

I think previous climbers had been very optimistic in reporting this candidate was visually the tallest,  because it is not. I’ve surveyed quite a few peaks around the world with multiple summit candidates of similar height, and I’ve learned it’s extremely challenging to make accurate assessments by eye. In this case it was unclear by eye which of the six candidates including the one I was on were tallest.

But I had the equipment to make an accurate assessment. I first pulled out a 2.5x zoom surveyors sight level and looked all around 360 degrees. I confirmed those five candidates were indeed the only ones close to my elevation. One due north was a summit with an overhanging boulder on the top that would be a challenging climb. The other four were in the area of the DEM candidates I’d identified on the Venezuela border to the West.

I then pulled out and assembled my tripod and mounted the 32x surveying auto level on top. I used the three dials to tune it so it was perfectly level, then independently pointed it at each of the summit candidates.

The overhanging boulder summit was very close in height, but still shorter. The angular resolution of my level is 20 arc seconds, meaning at a distance of 500m I can resolve a height difference of about 5 cm. Earth curvature correction at that distance is only a few cm. The working scope for the device is +/-15 arc minutes.

While the main purpose of the level in this application is simply to determine if a rival peak is shorter or taller, the level can also give a rough measure of angular declination or inclination.

The boulder was between 0.5-1 working scopes below horizontal, meaning an angular declination of 15-30 arc minutes. Based on satellite imagery the boulder peak was approximately 400m away horizontal, which, using basic trigonometry, works out to a 1.7m-3.5m height difference.

This height difference at a distance of 400m horizontal away is way too small to detect by eye in my experience, and would be within error bounds of a commercial drone altimeter. This means it cannot be definitively measured shorter by a commercial drone and cannot be asserted to be visually shorter.

I measured the other four candidates near the DEM locations and found them each shorter, around 0.5-1 working scopes angular declination. They were all close to 600m away horizontal based on satellite imagery, so were about 2.6m-5.2m shorter. This is also not possible to detect by eye at that distance in my experience, and within error bounds of a drone altimeter.

Based on these measurements, and given that I was at least 500m inside the Guyana border, I determined I was on the highest point in Guyana. The next goal was to measure an accurate elevation.

After verifying all rival points were shorter using the sight level and the auto level, I mounted the Trimble DA2 on a flexible leg tripod on the summit and started logging data at 11:15am.

I usually try for at least one hour of data because this is the minimum requirement for some processing algorithms. But on challenging summits where climbing time is more critical for safety I’ll go as short as a 30 minute measurement. That still generally gets errors down to 0.1m, which is pretty good for most peaks.

In this case I wanted the full hour, but after 30 minutes I saw rain clouds building back and the tepui getting shrouded in clouds. I knew the descent had now partially dried, but if it got rained on again it would be sketchier, especially the traverse part. So I called it good enough, and quickly logged the data and packed up.

We downclimbed the traverse, pulled the rope, then rapped the next two steps. We then scrambled to the chimney, which would be a tricky diagonal rappel. Andres offered that I could do a more direct single strand rappel to avoid the chimney, then he would go do the chimney rappel. That sounded fine with me, so we slung a horn and I did a free hanging rappel into the slot canyon and through the arch back to easy terrain. Then Andres pulled the rope up and rapped the chimney.

It started raining then, but at least we were done with the sketchiest parts. If I had waited longer on the top we would have been descending the sketchy parts in the rain. We then scrambled down the route in the rain, being extra carry on the slimey slabs.

We then did the short bushwhack and hike back to camp, rapped the last step, and were back by 2pm. We rested that afternoon as it rained off and on. At sunset in a brief clearing we hiked over to “The Window” which was a great view off the edge of the tepui to the Brazilian side below. Unlike the Venezuelan side, this side was all jungle.

July 1

I had budgeted two reserve days just in case the first candidate was not the true highpoint, so I could go climb the other candidates. But with the highpoint now established and measured, those days were not necessary.

So we hiked out in the morning in the fog and rain. We stopped at an interesting Crystal Valley near the triple country point, then continued back to Hotel Principal. Two other hikers from Italy and Brazil were staying  nearby at hotel Indio, but otherwise the area was empty.

That afternoon in a brief break in the rain I hiked over to the Jacuzzi, an interesting formation in a stream.

July 2

The next morning we hiked down the ramp off the tepui back to Rio Tek for lunch, then continued to Paraitepuy by sunset. We slept there that night, then returned to Santa Elena the next morning.

Discussion

I processed the results with TrimbleRTX and measured the Guyana highpoint is 2805.8m +/-0.1m EGM2008 elevation at location (5.210989, -60.739325). This location is squarely within Guyana, a distance 505m from the border, and I measured it was higher than all other candidate peaks in Guyana. Thus, that point is the country highpoint of Guyana.

This location is 51.6m taller than Maverick Rock, meaning it is the tallest point on the entire Mt Roraima tepui. Thus, this point is the true summit of Mt Roraima. So Mt Roraima is the country highpoint of Guyana, at 2805.8m, and the true summit is located in Guyana.

Any conclusion about the country highpoint location or elevation prior to these 2026 measurements would have been premature. Previous measurements were not taken with sufficient accuracy to determine which candidate highpoint was tallest, and elevation measurements had too high errors to be conclusive.

Interestingly, DEM elevations of the true summit of Mt Roraima were in error by 400m-500m vertical. These are the largest errors I’ve ever seen for satellite-based DEMs. It’s possible this is because Mt Roraima is a very unique shape, different than most mountains. It is mostly flat on top but towers stick up randomly and can be missed. The Labyrinth area where the summit is located is very complex terrain, with deep chasms next to narrow towers. Its possible measurements missed the towers, hit the chasms, and this led to extreme underestimates of elevations.

 

© 2026, egilbert@alum.mit.edu. All rights reserved.