The Plain of Jars (Lao: ທົ່ງໄຫຫີນ Thong Hai Hin, ) is a megalithic archaeological landscape in Laos. It consists of thousands of stone jars scattered around the upland valleys and the lower foothills of the central plain of the Xiangkhoang Plateau. The jars are arranged in clusters ranging in number from one to several hundred.
The Xiangkhoang Plateau is at the northern end of the Annamese Cordillera, the principal mountain range of Indochina. French researcher Madeleine Colani concluded in 1930 that the jars were associated with burial practices. Excavation by Lao and Japanese archaeologists in the intervening years has supported this interpretation with the discovery of human remains, burial goods and ceramics around the jars. Researchers (using optically stimulated luminescence) determined that the jars were put in place as early as 1240 to 660 BC. The jars at Site 1 (using detrital zircon geochronology) were determined to have been transported to their current location from a presumed quarry eight kilometers away. The Plain of Jars is one of the most important prehistoric sites in Southeast Asia.
Jar sites
Jars More than 120 jar sites have been identified in Xiangkhouang Province and surrounding areas, with 129 sites now confirmed by recent surveys. Today, UNESCO officially recognizes over 50 sites, but there are additional documented sites not officially recognized. Each site has between one and 400 stone jars. Documented sites now include over 2,100 jars. The jars vary in height and diameter between 1 m and 3 m and are all hewn from rock. Their shape is cylindrical, with the bottom always wider than the top.
Stone discs were also found. The discs differ from the lids, as they have at least one flat side. Previous theories that the discs could have been lids were disproved due to that none of the discs were found on top of a jar and none would fit on the tops of the jars. Fewer discs have been found than jars, totaling 200 discs. It is possible that the discs may have been used as burial markers, as the discs were found nearby jars or could have been placed on the ground to mark a burial pit. Some of the jars contained human remains. Evidence supporting that these discs were used as burial markers includes an extensive network of intentionally placed and largely unworked stones marking elaborate burial pits and chambers to the north of Xiangkhouang, known as the "standing stones of Huaphanh." These have been dated to the Bronze Age.
Iconography is found on many of the discs. These images consist of both anthropomorphic and zoomorphic representations. The anthropomorphic, or human-like, figures are more uniform than the zoomorphic, or animal-like, images. The anthropomorphic figures are almost all "spread-eagle" with straight or bent legs. Zoomorphic figures include tigers, monkeys, and other animals. Of the jars, only one jar has been found with iconography carved into it. This jar from Site 1 has a human "frogman" bas-relief carved on the exterior. Parallels between the "frogman" and the rock painting at Huashan in Guangxi, China, have been drawn. The Chinese paintings, which depict large full-frontal images of humans with arms raised and knees bent, date to 500 BC–200 CE.
The most investigated and visited Jar site is close to the town of Phonsavan, and is known as Site 1. Seven jar sites have now been cleared of unexploded bombs and are open to visitors. These are currently sites 1, 2 and 3, and Site 16 near the old capital Xieng Khouang; Site 23, near the big hot spring in Muang Kham; Site 25 in the largely unvisited Phou Kout District; and Site 52, the largest known jar site to date with 392 jars near a traditional Hmong village only accessible by foot.
Archaeological research The earliest documented archaeological research began with Madeleine Colani in 1935, who documented 21 sites and gave them each names. Colani documented additional possible sites, but later research often had trouble locating these sites due to differences in metric degree systems. The most well-known of the sites that Colani identified is Site 1, also known as Ban Ang. In Colani's initial excavation around 20 jars, spindle whorls, ear discs, glass beads, and bronze and iron tools, knives, and jewelry were found. Human remains and glass beads were found in some of the jars.
Colani also recorded and excavated at twelve Plain of Jars sites and published two volumes with her findings in 1935. Colani concluded that the Plain of Jars was an Iron Age burial site. Inside the jars she found, embedded in black organic soil, coloured glass beads and burnt teeth and bone fragments, sometimes from more than one individual. Around the stone jars, she found human bones, pottery fragments, iron and bronze objects, glass and stone beads, ceramic weights and charcoal. The bone and teeth inside the jars show signs of cremation, while the burials surrounding the jars yield unburnt secondary burial bones.
More recent excavations, such as those by the joint Lao-Australian Plain of Jars Archaeological Project (PJARP) from 2016-2020, have confirmed that Site 1 was used for secondary burials dating from the 8th to 13th centuries CE. These later excavations revealed that multiple individuals were interred in shared mortuary contexts. In 2023, preliminary excavations at Site 75 unearthed uncremated human remains inside a stone jar, the first substantive instance of such a burial in Lao jar contexts.
No further archaeological research was conducted until November 1994, when Professor Eiji Nitta of Kagoshima University and Lao archaeologist Thongsa Sayavongkhamdy surveyed and mapped Site 1. Nitta claimed that the surrounding burial pits were contemporaneous to the jars, as they were cut into the surface on which the jars had been placed. Nitta believed the jars were symbolic monuments to mark the surrounding burials. He dated the Plain of Jars to the late second or early first millennium BC based on the burial urn and associated grave goods. Sayavongkhamdy undertook surveys and excavations between 1994 and 1996, supported by the Australian National University. Sayavongkhamdy and Peter Bellwood interpreted the stone jars as a central person's primary or secondary burial, surrounded by secondary burials of family members. Archaeological data collected during bomb clearance operations supervised by UNESCO archaeologist Julie Van Den Bergh in 2004–2005 and again in 2007 provided similar archaeological results. Like Nitta, Van Den Bergh concluded that the jars and surrounding burials were contemporaneous.
Archaeological research conducted by Dougald O'Reilly, et al. in 2018 that covered 59 sites under the Safeguarding the Plain of Jars Project, or the SPJP's, report and dozens of undocumented and possible sites, revealed that jar sites were typically built in mountainous areas and most don't have many jars, with 16 sites having between 20-60 jars and only three sites with over 200 jars. Six rim styles were identified: flat rim, recessed inner rim, outer rim, recessed inner rim with an outer rim, prominent rim, and collar rim.
In 2026, archaeologists investigating Site 75 excavated a large stone vessel known as Jar 1 and recovered the remains of at least 37 individuals, including adults and children, dated between 890 and 1160 CE. The jar, measuring approximately 1.3 metres high and more than 2 metres wide, is the first example in the region to contain undisturbed human remains within the vessel itself. Analysis of the bones and teeth indicated that the jar was used repeatedly over several generations as part of a secondary burial practice in which human remains were deposited after earlier stages of mortuary treatment. Archaeologists also recovered glass beads and other associated materials, suggesting ritual or commemorative activity connected to the burial process. The discovery provides information for the funerary function of at least some of the megalithic stone jars scattered across the Plain of Jars and contributes to research on mortuary traditions in mainland Southeast Asia.
Jar geology The jars lie in clusters on the lower foothills and ridges of the hills surrounding the central plateau and upland valleys. Several quarry sites have been recorded, usually close to the jar sites. Five rock types have been identified: sandstone, granite, conglomerate, limestone and breccia. Quarry sites, such as those near Jar Sites 2 and 3, contain partially hewn jars still embedded in the bedrock, providing clear evidence of on-site extraction and initial shaping. It is assumed that Plain of Jars people used iron chisels to manufacture the jars, although no conclusive evidence for this exists. Experimental archaeology, however, has shown that lithic tools could also achieve the necessary shaping, and the jars could have been carved without the use of metal implements. Regional differences in jar shape have been noted. While the differences in most cases can be attributed to choice and manipulation of rock source, some differences in form (such as variations in the placement of jar apertures) appear to be unique to specific sites.
The majority of the jars are sandstone. The largest jars are made of sandstone and they are not found in plains or lowland areas. These large sandstone jars may only be found in more mountainous areas due to the more intensive techniques in quarrying and moving that were required for large jars, as opposed to small jars. A mountainous location would also allow for a close proximity to quarries. Analysis of detrital zircon from jars at Site 1 indicates the stone was transported from a quarry approximately 8 kilometers away.
Granite jars have only been found in one province, the Khun District, and are often found with a high degree of weathering. Many of the granite jars found have been very small, this might be due to the higher amount of time and effort required for cutting blocks from a granite rock mass.
Conglomerate jars are composed of a combination "medium gravel to cobble sized clasts set in a fine matrix," resulting in jars that are often found in very fragmented and poor condition.
Limestone jars have been found in two types: a weaker and a stronger variation. The weaker, honeycombed, brittle limestone exhibited more signs of weathering and were found in poor condition. The stronger, mid grey, reef limestone with fine-scale bedding exhibited less signs of weathering, but were still not found in the best condition. Not many sites have limestone jars, as they are only found at Site 39 (Ban Phonekham), Ban Tha, Site 41 (Phou En Kha), and Site 43 (Phu Hai Hin).
There is only one site where breccia jars are found: Ban Nom Hom. The site is located on a large-scale fault that resulted in a steep sided valley and hot springs found locally in the area that effected the composition and the separation of jars at the site.
Two principal iron ore deposits exist in Laos, both in Xiangkhouang. The presence and locations of the numerous jar sites in Xiangkhouang may relate to trading and mining activities. History has shown that Xiangkhouang, at the northern end of the Annamite Range, provides relative easy passage from the north and east to the south and west. Within the geographic setting of Xiangkhouang, the jar sites may reflect a network of intercultural villages, whereby the locations of the jars are associated with long-distance overland routes which connect the Mekong basin and the Gulf of Tonkin System. The jar sites show superficial regional differences, such as jar form, material, and the number of jars per site, but all share common setting characteristics such as burial practices, elevated locations, and commanding views over the surrounding area.
Mortuary ritual