Sacsayhuaman · Peru · -13.5078, -71.9822 · PYR record in the living atlas
Pedro Cieza de León saw the walls above Cusco in the 1550s and wrote that the stones were laid and joined so finely that a rial, a Spanish coin, could not be put in between them. In the same passage he wrote down the staffing. The Inca, he says, ordered the provinces to send 20,000 men, roughly 4,000 of them to quarry, 6,000 to haul, and the rest to dig foundations and cut timber, all of it worked in rotation. Those numbers are a chronicler's, not a survey's, and they should be read that way. The joint is what everyone photographs. The labour roster is what explains it.
How you fit a boulder
The polygonal joints look impossible and are merely laborious. The technique has a name worth using: polygonal, or cellular, ashlar, in which each stone is dressed on several faces to fit the particular neighbours it is going to meet. That is a different craft from the coursed rectangular blocks the Inca used in most of their other building. The tool marks say the blocks were pounded into shape with harder stone rather than sawn, and hammerstone chips turn up in excavation below Inca walls. Many blocks still carry the nodes and bosses left protruding to take rope slings. Three materials do most of the work in the Cusco region: Yucay limestone, the green diorite porphyry of the Sacsayhuaman hill itself, and black andesite. Recent LiDAR has tentatively identified quarrying on Sacsayhuaman's own slopes, which would put the source of the largest limestone blocks within a few kilometres of where they stand; the andesite used in some sectors came from further afield. Walls lean inward as they rise, typically by about five degrees, which is why Inca doorways and niches are trapezoids rather than rectangles.
The fitting was iterative. A block was lowered against its neighbour, the contact line marked, the block lifted, the high spots dressed away, and the whole cycle repeated until the two faces met. Scribing and fitting, the method a joiner uses, at the scale of houses. Experiments have shown that this laying, removing, re-cutting and re-laying was considerably quicker than scholars used to assume, which matters: the argument that the walls are too precise to be handmade rests almost entirely on an intuition about how long the work would take.
Jean-Pierre Protzen documented the whole industry in the field in the Urubamba valley at Ollantaytambo, where the quarries at Kachiqhata lie about 5 kilometres from the site across the river. He traced roads, ramps and slides running from quarry to wall, found carefully built roads of clay and gravel under abandoned megaliths, wooden rollers beneath another, and the piedras cansadas, the tired stones, dropped where the haulage stopped. Sacsayhuaman has the same signature: quarry faces, drag routes, and at the rear of the complex, in the area called Suchuna, terraces, patios, cisterns and aqueducts belonging to the same building campaign.
There was also a structural payoff. Interlocking blocks and inward-leaning faces make Inca walls very hard to shake apart. The earthquake of 1650 destroyed almost all of colonial Cusco. The quake of 21 May 1950 damaged more than a third of the city's structures, and the Inca masonry underneath came through it; the repairs that followed stripped away colonial superstructure and exposed Inca walls that had been hidden inside later buildings.
The weight nobody has weighed
Published estimates for the largest limestone block at Sacsayhuaman disagree badly with each other. Some accounts give about 128 tonnes, some almost 200, and recent summaries put the largest at somewhere between 70 tonnes and a little over 100 while saying outright that the figures are disputed. That spread is the honest headline. No one has put one of these blocks on a scale; every figure is a volume estimate multiplied by an assumed density, and small errors in either compound. Cieza, who had the advantage of seeing the walls before they were robbed, simply measured one stone at 260 of his palmos in circuit and left it there.
The overall dimensions are contested too. One published description gives the terrace walls as roughly 400 metres long and about 6 metres high, enclosing over 6,000 cubic metres of stone. Another gives more than 540 metres of zigzag in as many as 40 segments, with the three tiers reaching 18 metres, and individual megaliths standing to about 8.5 metres. We quote them rather than pick, because they are not measuring the same thing and neither is easy to verify from a photograph.
The claim that the Inca did not build it
The standing claim about Sacsayhuaman is that the megalithic lower courses are too fine for stone tools and therefore predate the Inca. Brien Foerster is the best known proponent of the two-phase version, in which the lower courses belong to an earlier and unnamed people and the smaller upper work is Inca. There is a modest version of the claim which is supported, and a large version which is not.
The supported version: the hill was not empty when the Inca arrived. Pottery studies put the earliest occupation of the hilltop at about 900 AD. The Killke culture held the region from roughly 900 to 1200 AD, and radiocarbon has been read as placing a fortification here around 1100 AD, well before the Inca expansion of the 15th century. On 13 March 2008 archaeologists announced further ruins at the edge of the complex, apparently ceremonial and probably Killke, with no agreed function. Recent LiDAR and ground-penetrating radar suggest the ridge was being shaped and modified well before Pachacuti's building programme, which leaves open the possibility of an earlier ceremonial centre whose character has not been established. So there is genuinely pre-Inca building on this hill, it is not a secret, and how far back it runs is an open question under active investigation rather than a settled date.
The large version, that a vanished people of unknown date raised the megalithic courses and the Inca merely inherited them, has to work around the documentary record. Cieza names the rulers under whom the work proceeded and describes the labour draft while men who had watched the empire fall were still alive. Pedro Pizarro inventoried the interior after the conquest and found rooms filled with lances, arrows, darts, clubs, bucklers and shields, which is a functioning Inca arsenal, not a ruin someone squatted in. And the same masonry, with the same tool marks and the same abandoned blocks on the same kind of drag road, occurs across the empire at sites nobody disputes. The ceramic sequences, the colonial documentation and the consistent Inca polygonal vocabulary at confirmed Inca sites all point the same way.
Fortress is a Spanish word for it
The imperial building campaign began under Pachacuti, whose reign is usually given as about 1438 to 1471, and continued under Tupac Yupanqui and Huayna Capac until roughly 1527. What the complex was actually for is not settled. The word fortress comes from the Spanish chronicles, written by men who had just fought a battle here, and modern researchers regard it as too narrow a label: the esplanade, the terraces, the cisterns and the ceremonial calendar all describe something more than a garrison. Fortress or temple is a choice the sources have handed down rather than one the evidence forces, and both readings have support.
What the Spanish took
The siege began on 6 May 1536 and ran into March 1537. Manco Inca's forces held the hill; the Spanish, roughly 190 men, 80 of them mounted, with several thousand indigenous auxiliaries, took the terrace walls at night with ladders. Juan Pizarro was struck in the head by a stone during the assault and died of it days later. Reported Inca numbers range from 40,000 to 200,000, a spread that tells you the chroniclers were estimating a crowd.
Then came the quarrying. Garcilaso de la Vega, born in Cusco, wrote that the Spanish pulled down all the smooth masonry in the walls, and that there was not a house in the city that had not been made of this stone. What survives is what was too heavy to be worth carting away, plus foundations. That is the whole answer to the obvious question of why the biggest stones are the ones still standing: they were uneconomic to move. Spanish eyewitnesses describe a large round tower of four or five storeys standing inside the complex. That tower is Muyuqmarka, and its circular foundation on the summit is the clearest surviving trace of the upper architecture, alongside the footprints of two other towers, rectangular halls and the esplanade. Spanish authorities also had parts of the ruins covered with earth to keep them from being used again by rebel forces.
The ground is still being read
Three separate programmes have reported recently, and between them they are changing both the size and the colour of the place.
A joint season between June and August 2024, run by Peru's Ministry of Culture Cusco Directorate and the University of Warsaw under Sabino Quispe and Alexei Vranich, recovered a fragment of an Inca mace, a large assemblage of human, monkey and jaguar teeth with many of them perforated as if for ornaments, and traces of white, red, blue and yellow pigment that once decorated floors and walls. That last item ought to change how the site is described and illustrated. These surfaces were painted. The bare grey stone everybody photographs is what four and a half centuries of weather left behind.
A drone LiDAR and ground-penetrating radar survey running from 2024 to 2026, by the University of Warsaw with Italy's CNR ISPC, has covered roughly 69 hectares of the site and its slopes and found previously unrecorded terracing, roads, structures and possible quarry areas beneath dense vegetation. The known extent of the complex is larger than it was, its relationship to the surrounding terraces and roads is only now being mapped systematically, and the work is not finished.
And the chinkana. Local tradition holds that passages once ran from Sacsayhuaman to the Qorikancha in central Cusco, and the treasure stories attached to them have no support of any kind. In early 2025 the Peruvian archaeologists Jorge Calero Flores and Mildred Fernández Palomino announced ground-penetrating radar evidence for a tunnel system and began targeted excavation. Nothing is published in peer-reviewed form yet, and the date, function and extent of any passages are all unresolved. What has changed is the status of the question. It is archaeology now rather than folklore, and that is worth saying without pretending the answer is in.
The head of the puma, and 24 June
The description of Cusco as a puma with Sacsayhuaman for its head is not a modern invention: it comes from the chronicles, from Juan de Betanzos and Pedro Sarmiento de Gamboa, in which Tupac Inca relates that his father called Cusco the lion city, with the tail at the meeting of the rivers and the body at the great square, and that the head was wanting, so a fortress was built on the high ground to the north. Whether the plan was laid out literally as an animal, and how the largest stones were quarried and moved, are both still undetermined in the published record. That is an unresolved question, not a suppressed one.
The ceremony did come back. Inti Raymi, the solstice festival, was last celebrated with the emperor present in 1535 and was then suppressed. In 1944 Faustino Espinoza Navarro staged a reconstruction built from Garcilaso's chronicle, with indigenous performers, and since then it has been held at Sacsayhuaman every 24 June. The script is a 20th century reading of a 17th century book about a 15th century rite, which is worth knowing while you watch it.
What our engines found
We ran the same test here as everywhere else, with the settings fixed in advance and identical at every one of the fifty sites we cover: a 200 km search radius, at least 12 km between sampled sites, 15 sites, a 2 per cent tolerance, and 2,000 Monte Carlo runs against a null model resampled from real catalogued sites in the same region. The settings are frozen because the answer moves if you tune them, and tuning them after seeing the answer is exactly the trick we object to elsewhere. Sacsayhuaman returned 50 golden-ratio pairs where chance predicts 32.3, giving p = 0.0380.
That clears the usual significance line, and the control is why we are not going to celebrate it. Alongside phi we test arbitrary ratios carrying no mystical claim whatsoever, and here the best of them scored 1.52 times its expected count against phi's 1.55. That is a tie. A result a meaningless number can match is telling you about the density of monuments in this region, not about the golden ratio.
We publish the misses because a tool that only reports its hits is not measuring anything. Every figure above is reproducible from the same coordinates, the same settings and the same seed.
One limit worth stating plainly, because it is the one most often glossed over. This test compares positions between monuments. It says nothing about any building's own proportions, its height, base, slope or perimeter. That is a separate question, and at several sites a genuinely open one.
Explore it yourself
Open the Sacsayhuaman record in the living atlas to walk the terraces in 3D, see the site among its neighbours in the Cusco valley, and run the alignment and ley line engines on the real coordinates. The Sacsayhuaman site-emblem tee carries the zigzag wall line, the site name and the exact coordinates in the Pyramason Chronicle style.
Sources worth reading
Protzen, J.-P., "Inca Stonemasonry", Scientific American (1986). Protzen, Inca Architecture and Construction at Ollantaytambo, Oxford University Press (1993). Hemming, J., The Conquest of the Incas (1970), for the siege of 1536. D'Altroy, T. N., The Incas (2002). CNR ISPC and the University of Warsaw, "Sacsayhuamán emerges from the forest", on the 2024 to 2026 LiDAR and radar survey.