The Masrur Rock-Cut Temples in Himachal Pradesh Were Carved From One Sandstone Ridge About 2,500 Feet Up, in a Block Some 160 Feet Long, and Scholars Date Them From the 8th to the Later 9th Century

At Masrur, a hamlet in Kangra district of Himachal Pradesh, craftsmen cut a complex of temples out of the crest of a sandstone ridge about 2,500 feet above sea level. H. Hargreaves of the Archaeological Survey of India (ASI), who published the first survey in the report for 1915 to 1916, measured the carved block at some 160 feet long and at least 105 feet wide. He called it unique in the Himalayan region and rare in northern India. No inscription gives a date, and scholars place the work between the early 8th century and the later 9th.

The complex is often called the Ellora of the North, a name both Sahapedia and the Ministry of Tourism’s Incredible India site use. Hargreaves set it beside the three groups of free-standing rock-cut temples then known: the rathas at Mamallapuram (see Mahabalipuram, the Pallava shore), the Kailasa at Ellora (see the Kailasa temple at Ellora) and the temple at Dhamnar. He wrote that Dhamnar and the Kailasa lie in pit-like hollows while Masrur stands on the highest point of its ridge, and that Masrur is three times the length of its rival. Dhamnar is in Mandsaur district of Madhya Pradesh, according to Puratattva and a 2024 journal paper; Hargreaves, writing in 1915, placed it in Rajputana.

A temple cut down from the top of a ridge

Two roughly parallel cuttings, each 11 to 18 feet wide, separate the central block from the rest of the outcrop, Hargreaves reports, and the block between them was carved into shrines. The main shrine faces a little north of east, toward the Dhauladhar range, with smaller shrines stepping down in size on each side and at the back. In front of the portico stand two free-standing, cross-shaped temples, and in front of everything lies a tank 155 feet by 85 feet that Hargreaves says is reputed to hold water even in the driest seasons.

The stone varies from place to place. In Hargreaves’s account, some carving looks as if it had left the sculptor’s hand the day before, and other stretches have weathered to faint outlines. Where the rock was too crumbly, the craftsmen set in better stone. Only the main shrine and its adjuncts were fully excavated, and several doorways were begun but not finished. Michael Meister, in a 2006 article, also calls the temple never completed.

Fifteen spires, or nineteen temples

Modern sources disagree on the count. Incredible India says 15 temples survive, and Sahapedia says 19. Neither says how it counted, but the two earliest accounts point to a likely source. Thirteen shrines cut from the same rock plus the two cross-shaped temples make 15 in Hargreaves’s list. Shuttleworth’s plan of 1915 numbers 15 spires and four half-spires, which makes 19 numbered parts. Shuttleworth also counted probably 28 doorways or porches at ground level.

Who recorded it, and when

The first traceable mention is in a list of antiquarian monuments published at Lahore in 1875. It names a Thakurdwara temple in the village of Masrur, “in good order”, not photographed and said to date from the time of the giants. A revised list of 1891 says the ridge had been cut through in two places and carries no inscriptions.

H. L. Shuttleworth, a civil officer then at Hoshiarpur, first visited in April 1913 and returned in November to draw a rough plan. Hargreaves examined the site in October 1913, a visit Shuttleworth says his photographs had largely prompted. Shuttleworth’s note in The Indian Antiquary of January 1915 says the carving is of a standard that no other early temple in those parts equals.

The Government of the Punjab made the excavation and conservation possible. Hargreaves reports that the front of the temple lay under about eight feet of earth and that a cowshed, huts and fields had grown up around it. These were cleared, and the priest was moved to new buildings on a purchased site. The report was printed in Calcutta in 1918.

Eighth century or later ninth

With no inscriptions, Hargreaves wrote, only style could guide the dating, and by that guide the monument could hardly be earlier than the 8th century and might be somewhat later. Others have since narrowed or shifted that range.

• Meister (2006) says the temple was cut in the first half of the 8th century, and that its creation marked a movement of political power into the hills. He also argues that Masrur is the conceptual model for the temple-mountains Cambodian kings built later.
• Krishna Deva, a former ASI Director General, wrote in 1969 that the complex belongs to the later half of the 9th century and to the Pratihara period, as quoted by Sahapedia.
• N. K. Singh, whose book the Tribune names as Coronation of Shiva: Rediscovering Masrur Temple, argues for the 6th to 8th century and suggests the king was a devotee of Shiva, most likely Yashovarman. The Tribune says he regards the question as still open.
• Incredible India gives the 8th to 9th century, and the Kangra district page gives the early 8th century.

On who commissioned it, Sahapedia cites O. C. Handa’s view that local rulers did not have the means, and that wealthy traders may have paid migrant artisans from Kashmir. Sahapedia notes that some historians credit the Katoch rulers while others say the dates do not fit. It also records, citing Aurel Stein, that the Kangra region came under the king of Kashmir in the 8th century. For that court, see Lalitaditya Muktapida and the Martand sun temple.

Shiva, Vishnu or Rama

The main shrine is known locally as the Thakurdwara, which Hargreaves glosses as the shrine of Vishnu. Its sanctum holds three stone images identified as Rama, Lakshmana and Sita, and Incredible India says the main shrine houses them today. These were recent additions, Hargreaves wrote: twenty-six years before his report only a small metal image was worshipped there. Shuttleworth saw the three black stone images in 1913.

The carving points elsewhere. On the lintel of the main shrine, Hargreaves reads nine seated figures with Shiva at the centre, among them Vishnu, Indra, Ganesha, Skanda and Durga. This he took as a strong presumption that the temple was first dedicated to Mahadeva, a name of Shiva. Niches elsewhere hold three images of Shiva, two each of Indra and Surya, and one each of Kartikeya and Ganesha. Hargreaves found Vishnu identifiable with certainty only once, and Shuttleworth thought the question of dedication hard to settle without more excavation.

The Pandavas and the unfinished night

Local tradition holds that the Pandava brothers, in exile, cut the temple in a single night, which Hargreaves says the tradition made six months long. The accounts differ on why the work stopped. Hargreaves was told that a woman busy at an oil press lit her lamp, went out to see why the night would not end, and so interrupted the work. Shuttleworth was told of a telin, a woman of the oil-pressing community, who came out of her house just before dawn. Incredible India and the 2024 journal paper tell instead of a staircase to heaven that the Pandavas vowed to finish before daybreak, until Indra raised a bird’s call before its time. Two unfinished staircases are still part of the complex, the paper notes.

Weather, earthquake and the ASI

The sources disagree about the great Kangra earthquake of 1905. O. C. Handa, quoted by Sahapedia, writes that it passed over the temple “without a scratch on its monolithic solidity”. Meister’s abstract says successive earthquakes sheared the stone and folded parts of the complex back into the hill. Incredible India, the district administration, the Tribune and a Himachal Headlines report of 3 November 2026 also name the earthquake as a cause of damage. Hargreaves’s report, written a decade after 1905, blames weathering, rock falling away along its natural cleavage, the roots of pipal trees and the use of the shrines as stores and cattle sheds, and its text, searched in full, does not mention the earthquake.

The ASI’s list of centrally protected monuments carries the Masrur monolithic temples, with sculptures lying in and outside them, under the Shimla Circle in Kangra district. The Tribune reported in 2019 that the sandstone is prone to weathering, that the ASI had restored many portions, and that unnamed sources said a hardening treatment had proved only partly effective.

Getting there

As checked on 5 November 2026, Incredible India lists opening hours of 7:00 am to 7:00 pm. The Kangra district page gives road distances of 41.3 km from Dharamshala, 226 km from Shimla, 234 km from Chandigarh and 253 km from Manali. It names Gaggal airport as the nearest, and Daulatpur Chowk, on the Pathankot to Jogindernagar narrow-gauge line, as the nearest railway station. Incredible India instead lists the Pathankot, Chandigarh and Amritsar airports and Pathankot station. The Tribune put the temples about 30 km from Dharamsala in 2019. The two official pages differ, so confirm the route and hours before setting out. For the state, see Himachal Pradesh at a glance.

On 2 November 2026 the Governor of Himachal Pradesh visited the complex and was briefed on its architecture and conservation work, Himachal Headlines reported the next day.