How long did the construction of the Giza pyramids take?

The standard answer is: about twenty years. It appears in encyclopedias, on signs, and in practically every depiction, and it is likely approximately correct.

Remarkable is what it is based on. There is no Egyptian source that names a construction time. No cornerstone, no dedication inscription, no administrative note with start and end dates. The number comes from a Greek historian who wrote two thousand years after the events and relied on stories told by Egyptian priests.

Nevertheless, it holds up, not out of convenience. It fits with several independent boundary conditions, and no better number has emerged so far. This is a weaker foundation than most assume, and a stronger one than it first appears.

This page shows where the twenty years come from, what calculations are usually made with them, and at which point these calculations make a crucial error. For the most famous one—the block every two minutes—conflates two entirely different transport chains.

First, a clarification that simplifies everything. When people talk about construction time, almost all figures refer solely to the Great Pyramid of Khufu. The three great pyramids of Giza were built sequentially under three different kings, and roughly eighty years separate the first and last structures. Thus, there is no single construction time for the entire site, but three separate projects that overlapped slightly at the edges.

Something else is important: Construction time is not a measured quantity, but an estimate. There is no document recording it, nor a method to read it directly from the stone. What we have are boundary conditions. From these emerges a corridor, and the twenty years lie roughly in the middle of this corridor.

Where the twenty years come from

The source for the twenty years is Herodotus, and it is worth reading exactly what he writes.

He reports in his second book that the construction took twenty years, that one hundred thousand men worked in shifts of three months, and that an additional ten years were needed for the causeway and preparatory work. Strictly speaking, he thus names thirty years for the total project and twenty for the structure itself.

One of these statements is certainly false. One hundred thousand people simultaneously is ruled out by excavation findings; the settlement at the site was designed for a few thousand. The number is likely an exaggeration, as often occurs in ancient historiography, or a statement about the total number of all people ever involved.

The shift duration of three months, however, seems surprisingly credible today because it corresponds exactly to the length of the Nile flood, during which the rural population was not needed on the fields. Herodotus thus transmitted a real mechanism, even if he missed the magnitude.

Regarding the construction time itself, there is no comparable counterargument. It stands because nothing contradicts it.

Still, one should consider how this information reached Herodotus. He traveled through Egypt in the fifth century BC and had priests tell him what stood in their traditions. Between the construction and this conversation lies around two thousand years, roughly the same gap that separates us today from Roman times. What he wrote down is therefore not a source from the construction era, but a very late report about a very old memory.

That precisely the number twenty survived while the names of the builders were lost says something about the nature of this tradition. Round numbers hold better than details. It is therefore quite possible that twenty years was never an exact figure, but from the beginning an idiom for a very long construction project.

Longer than the king reigned, it could not go

There is a hard upper limit, and it comes from a completely different direction: A king could not build longer than he reigned.

A pyramid was a tomb. It had to be finished when its owner died, and after his death, it was at best hastily completed. Construction time is thus tied to the reign duration, and for Khufu, the transmitted figures range between just over twenty and clearly over thirty years, depending on how one reads a specific administrative record.

This results in a window rather than a single number. Twenty years are compatible with any of these readings; thirty only with the longer one. Whoever estimates the construction time upward must also assume a longer reign, and both assumptions rest on the same uncertain foundation.

From this coupling follows something else, rarely mentioned. Construction must have begun early in the reign, practically immediately upon accession. A king who waited ten years risked ending up in an unfinished tomb, and exactly this case is documented for other pyramids.

Thus, construction time was not a schedule drawn up, but a race against a deadline whose date no one knew.

This deadline also explains why certain structural elements were created so early. The underground chamber and the lowest courses were completed first, and several experts suspect that the burial complex was redesigned in stages: first a tomb deep in the rock, then a chamber in the masonry, finally the current chamber high up. Each of these changes makes sense if a king realizes he has more time than initially thought.

Conversely, this means: The finished pyramid is not necessarily the original plan. It is what fit within the available time, and the available time was the only variable no one could control.

The calculation with 2.3 million blocks

The most famous calculation on this topic goes like this, and it appears in half of all documentaries.

One takes around 2.3 million stone blocks and distributes them over twenty years. This yields about 315 blocks per day if calculated without breaks. With a ten-hour workday, that is well over thirty blocks per hour, or roughly one block every two minutes, for two decades.

This number is mathematically correct and is also used as a starting point in current academic literature, such as in a study on counterweight systems in construction from 2025. It is the reason why many people find the construction time implausible, and it is simultaneously the reason why some conclude there must have been unknown technology.

Before jumping to that conclusion, it is worth looking at what is actually counted in this calculation. One block every two minutes does not mean that a block was moved every two minutes. It means that, on average, one block reached its final position every two minutes, distributed across a construction site with sides hundreds of meters long, where many columns worked simultaneously.

With twenty working teams in parallel, this corresponds to one block per team every forty minutes. This is no longer a superhuman feat, but a matter of organization.

Furthermore, the blocks are distributed very unevenly. The lowest courses are the thickest and contain by far the largest part of the mass; towards the top, the stones become smaller and the courses thinner. Around half of the total volume is contained in the lowest fifth of the height. Thus, the material throughput was colossal at the beginning and decreased steadily afterwards, while transport routes upwards became longer and more difficult.

Whoever distributes the blocks evenly over twenty years therefore describes not a single real week on this construction site. The average is mathematically correct but describes a construction process that never existed.

The error in almost all these calculations

Now to the error contained in almost all these calculations, and it is not a rounding problem, but a logical one.

The number of 2.3 million blocks describes the entire structure. The idea that all this material was brought via the Nile, however, is false. Around ninety percent of the mass consists of coarse limestone, sourced from quarries a few hundred meters south of the construction site. This material never needed to be loaded onto a boat, was not dependent on water levels, and could be moved year-round.

Water transport was used for only two materials: the fine white limestone for the outer casing, from Tura on the opposite side of the Nile, and the granite for the chambers, from Aswan. Together, these constitute a small proportion in quantity and represent the actual logistical challenge, as they depended on ships, water levels, and seasons.

Thus, two completely different chains run alongside each other. One is a mass transport over short distances, uniform and year-round. The other is a specialized logistics operation over hundreds of kilometers, seasonal and weather-dependent.

Whoever compresses both into a single number gets a calculation that fits neither. And exactly this happens when one distributes 2.3 million blocks evenly over twenty years.

The difference has concrete consequences for the construction flow. Core masonry could be prioritized, stacked, and temporarily stored if necessary; it came from next door and was arbitrarily available. Cladding, on the other hand, had to be present in the correct sequence and at the right time, because it was set from top to bottom and delivery was tied to water levels. A missed river flood meant a lost year for that specific material.

Therefore, the question of construction time is fundamentally a question of cladding. The rough structure was a quantity problem. The outer skin was a scheduling problem, and schedules are what construction projects fail on.

What the Diary of Merer reveals

We have known fairly precisely since 2013 how the second chain actually functioned, and the numbers are soberingly concrete.

The service diary of the inspector Merer, found in harbor galleries at the Red Sea, documents a period of about five months in the 27th regnal year of Khufu. Merer led a crew of around two hundred men, bringing limestone blocks from the quarries of Tura to Giza, over a distance of roughly fifteen to twenty kilometers.

The rhythm can be read from his entries. About every ten days, the crew completed two to three round trips. The loaded journey took two days, the empty return trip one. Per month, this amounts to around two hundred blocks transported by a single crew.

Crucial is the period covered by the service diary: from the flood season into early spring, i.e., roughly July to November. Exactly these months saw the transport of heavy loads, because the river was then high and closer to the construction site. The evaluation of these records shows an operation oriented by water level, not by a construction schedule.

For the construction time, this means: Cladding could only be delivered during part of the year. The core masonry, however, continued to grow.

Noteworthy about Merer's records is also how casual they sound. He notes journeys, cargoes, and stays in a tone that reveals no state of emergency. It reads like a logbook, not a report on a century project. Precisely this is why these texts are so valuable: They show construction as routine business.

And they date themselves. The entries name the 27th regnal year of Khufu, and at this time, cladding material was still being delivered. If one estimates the reign at just over twenty years, this presents a problem, because in the 27th year, construction was obviously still ongoing. The records thus speak more for the longer reading of the reign duration and therefore for a construction project that spanned more than twenty years.

The years before the first block

What does not appear in any of these calculations is the time before the first block, and it was considerable.

Before construction could begin, the bedrock had to be leveled, with centimeter-level accuracy over an area of more than five hectares. Alignment to true north had to be surveyed without a compass and without the possibility of later correction. Quarries had to be opened, ramps laid out, harbor basins excavated, and channels dug.

In parallel, infrastructure for the people arose: housing, industrial-scale bakeries, breweries, storage, workshops, and administration. The excavations at this settlement show a complex that itself represents a major construction project.

Herodotus counts ten years for these preparatory works, in addition to the twenty for the pyramid. Whether the number is accurate is open, but the order of magnitude is plausible. Whoever states the construction time of a pyramid almost always means only the main phase and leaves out preparation and follow-up work.

More honest would therefore be a formulation distinguishing between the two: around twenty years for the structure, and a total project that ran considerably longer.

The same applies to the end. With the last cladding stone, the complex was not finished, but only the main construction. Afterwards followed the funerary temple to the east, the causeway down to the valley, the valley temple by the water, the subsidiary pyramids for the queens, and the pits for the disassembled boats. Each of these structures was, in itself, a larger project than what many cultures have ever erected.

When you stand on the plateau today, you see therefore a slice. Remains of the causeway are preserved, almost nothing of the valley temple remains under today's development, and the subsidiary pyramids look like footnotes next to their neighbor. But construction time applies to the total work, not just the silhouette one photographs.

What the neighboring pyramids reveal about time pressure

A look at the neighboring buildings shows what happened when the calculation did not add up.

The Menkaure Pyramid, the third and smallest in Giza, has a base area of granite whose blocks are only roughly dressed. Smoothing was apparently executed from bottom to top and abandoned midway through the work. The patron died before the cladding was finished, and what happened afterwards sufficed only for a closure, not for completion.

The same pattern is found elsewhere. Several pyramids of the Old Kingdom remained unfinished, some were not continued after a change of reign. This is the best existing evidence that construction time was indeed tied to a person and not to a construction program.

For the Great Pyramid of Khufu, this means conversely: It was finished. Cladding, chambers, sealing stones, everything executed. Whoever estimates a construction time must therefore assume a duration that fit into a reign and allowed for buffer time for delays.

This is a stronger argument for twenty years than any calculation with blocks per minute. Not because twenty years are proven, but because the structure is completed and the reign was not unlimited.

The middle pyramid adds another hint to the picture. At its peak sits to this day a remnant of the original cladding, and below it, it is almost completely missing. The reason lies not in construction time, but in later use as a quarry: What was easy to reach was removed, what sat high up remained. Questions of construction time and questions of preservation are often confused here.

This is useful for classification. Missing cladding does not automatically mean unfinished, and rough stone does not automatically mean abandoned. Only when processing marks end abruptly in the middle of a surface, as at the base of the third pyramid, is there truly an abandonment.

How research today handles the question

How contemporary research handles the question differs significantly from popular representation.

In academic literature, construction time is rarely the topic. Procedures, material flows, and labor organization are discussed, and the twenty years run along as a working assumption because they fit the boundary conditions. Nobody seriously tries to prove them, and nobody considers them disproven.

Practical experiments have nevertheless changed the discussion. Reconstruction attempts with contemporary tools show how long individual steps actually take, thereby providing lower limits: Striking a block from the rock, dressing it, moving it over a certain distance. These numbers are more reliable than any extrapolation from the total volume, because they are measured and not derived.

What they show is unspectacular and important: No single work step is impossible with the means of the time. The bottleneck was never an unknown technology, but the quantity and coordination.

That the entire necropolis has been protected as a World Heritage Site since 1979 has, incidentally, a practical influence on this research: Experiments on the original are excluded, hence reconstructions take place next to the site.

A second development is shifting the question. New investigations into the course of a long-silted Nile branch show that the water reached much closer to the plateau during construction than today and that its level fluctuated significantly over the centuries. A high level shortened the land route for heavy deliveries considerably; a low level extended it.

Thus, construction time becomes a variable that also depended on climate. A consequence of dry years could have slowed the delivery of cladding stones without anything going wrong in the construction itself. This dependency is not contained in any of the popular calculations, and it is probably the most important uncertainty factor overall.

Frequently asked questions

Where does the figure of twenty years come from?

From Herodotus, who wrote two thousand years after the construction. There is no Egyptian source with start and end dates.

Is the number therefore worthless?

No. It fits the transmitted reign duration, the size of the workers' settlement, and the delivery frequency documented by the service diary of Merer. Nothing has contradicted it so far.

Was construction carried out year-round?

For the core masonry, presumably yes, because the material came from the plateau itself. Transport of cladding and granite was tied to high water levels and occurred only during part of the year.

Does preparation count?

In the usual figure, no. Leveling, aligning, quarries, and the workers' settlement came beforehand, and Herodotus counts another ten years for this.

The short answer, and the robust one

If someone asks how long the construction of the Great Pyramid of Khufu took, the best answer is therefore twofold.

For the structure itself: around twenty years, a second-hand figure that fits all known boundary conditions and is contradicted by nothing. For the total project including preparation, temples, causeway, and ancillary facilities: considerably longer, presumably the greater part of a reign.

But the more interesting part of the answer is not the number, but what it presupposes. Twenty years of uninterrupted construction activity mean twenty harvests, during which enough surplus was generated to feed tens of thousands. It means an administration that functioned for two decades, across personnel changes. And it means that nothing major went wrong during these twenty years: no famine, no coup, no lost border.

This is the real achievement, and it stands in no calculation with blocks per minute. A structure of this size is less a technical than a political result.

And perhaps this is why the question of construction time is asked so persistently. It sounds like technology, and the answer is about statecraft.

Whoever wants to realize this on site has it easier than it sounds. The quarries south of the Great Pyramid are open, the quarry edges are visible in the rock, and the short distance to the construction site becomes immediately comprehensible. After that, the view of the base of the third pyramid, where work breaks off mid-stone. Between these two points lies the whole story: unlimited material on one side, limited time on the other.

The twenty years are ultimately not a measurement, but the best available estimate. It comes from second hand, it fits everything else we know, and it will likely not change significantly. What changes is our image of what actually had to be organized during these twenty years.