Overview
Dwarka (Sanskrit Dvārakā) is a pilgrimage city on the Gomati creek on the Gujarat coast, in the present district of Devbhumi Dwarka. It carries three overlapping sacred identities: one of the four Char Dham, the circuit whose other stations are Badrinath, Puri and Rameswaram; one of the Divya Desam, the shrines hymned by the āḻvārs; and a member of the Sapta Purī, the seven cities whose pilgrimage is held to confer mokṣa [S3][S5]. This entry concerns the city and its grouping, not the god — the presiding deity is Dvārakādhīśa, "Lord of Dvārakā", a form of Kṛṣṇa, who has his own entry in dwaraka-kalyana-narayana.md.
The status that needs the most care is the third. "Sapta Mokṣapurī" is not a fixed canonical list but a traditional grouping whose membership varies by text and region, and the sources located for this entry do not agree on it. The variation is not merely regional: it appears to begin inside the verse that is the usual basis of the grouping (see Sacred Narrative). This entry therefore reports the variants with their sources rather than asserting a single list.
The Place
The town stands on the banks of the Gomti creek, and the main shrine sits on the creek itself; the district administration describes it as seeming to rise from the waters of the Arabian Sea [S9]. Dwarka railway station is about 2 km from the temple, and NH-947 runs directly to the town; the nearest airports are at Jamnagar and Porbandar [S9]. The city is the westernmost station of the Char Dham circuit, and the temple trust's own district history describes Okhamandal — the wider region of which Dwarka is the chief town — as having passed between Yadava, Kalā, Vagher, Vadhela and British control [S12].
The physical evidence does not support a submerged ancient city. Sankalia's excavation on the western side of the present Jagat Mandir concluded that the present Dwarka is not earlier than about 200 BC, and a Kharosthi inscription in the first floor of the Sabhamandapa is assignable to that century [S8]; the material from the 1963 excavation has been reported as only about two thousand years old [S15]. The underwater explorations are real, but the harbour site dates only to historical times, with some underwater structure late medieval, and coastal erosion was probably the cause of the destruction of what had been an ancient port [S7].
Sacred Narrative
Dvārakā is Krishna's city on the western coast in the Mahabharata [S17]. The tradition of the land being reclaimed from the sea, and of the city being later given back to it, is the temple's own account rather than an archaeological finding; the trust's history page states that after Krishna's death the city "submerged under the sea" [S12]. No source located in this pass dates that submersion, so none is given here.
The Śālagrāma episode is the most widely known local story. The article reached reports the tradition that the devotee Bodana carried the original image of Kṛṣṇa away to Dakor, and that the image was later found in the step well at Dwarka called Savitri vav, historically recorded as Shivadhan Vav; a tradition of Vallabhācārya re-establishing the idol at Lāḍvā in 1560 Vikrama also appears on the trust's own chronology [S12][S13]. These are reported as the traditions they are. The date at which the present icon was actually installed is a firmer historical claim: Aniruddhaśrama Śaṅkarācārya installed the present image of Dvārakādhīśa in 1559 [S10], and the trust's chronology independently gives 1559 for the making and installation of the image [S12].
Scriptural Basis
The usual basis for the seven is a single Sanskrit verse, given in the source reached as Ayodhyā Mathurā Mayā Kaśī Kañcī Avantikā / Purī Dvārāvatī caiva — sapta etāḥ mokṣadāyikāḥ [S2], "these seven are the givers of liberation". The English rendering attributed to Wood's 1911 translation runs: "Ayodhyā, Mathurā, Mayā, Kaśī, Kañchi, Avantikā, Dwārāvatī,— these seven cities should be known as the givers of liberation" [S1]. Neither the Sanskrit nor the translation was read in the original for this entry, and both are flagged in Known Limitations.
The verse is the pivot of the whole question, and it will not bear the weight usually placed on it. As printed, it names eight places — Ayodhyā, Mathurā, Mayā, Kaśī, Kañcī, Avantikā, Purī, Dvārāvatī — against the word sapta, seven. The English rendering attributed to Wood drops "Purī" so that the count works, and it is precisely this surplus that different editors have resolved in different ways. The three lists below are therefore not simply regional drift; they are three answers to an ambiguity in the transmitted text. That is a reading of the evidence, not a collation of a critical edition, and a Sanskritist should check it.
Worship Practice
The most distinctive feature of the site's ritual geography is its two gateways: pilgrims enter by the swarga dwar, the gate to the heavens, and leave by the moksha dwar, the gate to liberation [S9]. A flight of 56 steps on the Gomati side of the temple descends to the water [S9]. Bathing at the ghat at the foot of those steps is therefore a documented part of the pilgrimage. The further claim that bathing here confers mokṣa is a devotional belief that no source reached in this pass states, and it is not asserted here.
Kṛṣṇa Janmāṣṭamī, also called Gokulāṣṭamī, is the principal festival. The trust's own published 2025 calendar gives it as 16 August 2025, Sravana vada 8, with Parna nauṁ the following day, and also lists Rādhāṣṭamī, Gopāṣṭamī, Nṛsiṅha Jayantī, Dīpāvalī and the Annakūṭ offering [S11]. The trust states that its dates are subject to change, as the festivals are celebrated according to the Gujarati calendar, and republishes the calendar annually. Harshananda records that the Janmāṣṭamī observance at this shrine was commissioned by Vallabha (1473–1531), and gives the temple's opening hours as 6.00 a.m. to 1.00 p.m. and 5.00 p.m. to 9.30 p.m. [S16].
The flag above the shikhara is part of the temple's public identity: the district administration describes it as made from 52 yards of cloth and visible from 10 km away [S9]. The claim that it is changed four times a day, on which devotees pay, carries a citation-needed tag in the article reached and is therefore not repeated here.
Temple(s) / Seats
The shrine is the Shri Dwarkadhish Temple, also called the Jagat Mandir or Nija Mandir, dedicated to Kṛṣṇa as Dvārakādhīśa [S13]. The district administration describes a building of soft limestone consisting of a sanctum, a vestibule and a rectangular hall with porches on three sides, with a carved shikhara rising to 43 m [S9]. It gives the district Archaeology Department's opinion that the temple is 1200 years old, and records the local belief that Vajranabh, Krishna's great-grandson, built it more than 2500 years ago and that it was erected in a single night under his direction [S9].
The building on the site today is late. The original structure was sacked by Mahmud Begada in 1473, and the existing temple is dated to the sixteenth century [S13]. That sack had a motive beyond religion: Mahmud Begarha attacked Dwarka largely because it harboured pirates who ravaged traders, the occasion being a complaint brought by a maulana whose ship had been plundered [S9][S14]. Rangarajan's reading is more qualified, and is given here in preference to a flat "destroyed": the temple was probably demolished in the fourteenth century by Śamsuddīn Daṅgānī and again in the fifteenth by Mahmud I, and the Veghars "did their utmost to defend" it and restored it when necessary [S10]. The site's own history records a further break in 1241, when five Brahmin families are said to have been martyred defending it, and repairs in 1730, 1861 and 1958, with the Government of India taking over care of the temple from 1960 [S12].
Regional Forms
Bet Dwarka, the island off the coast, carries a Rukmini shrine and is a separate place of pilgrimage, not a sub-shrine of the Jagat Mandir. The image taken to Bet during a Turkic attack, and the image at Lāḍvā, belong to the same cycle of stories about the icon's journeys [S12][S13]. Dvārakā is also claimed for the Dakor cycle, where the Bodana story is the local account of how Kṛṣṇa came to that town; that tradition is treated in the Dakor entries rather than here.
Known Limitations
The list of seven is genuinely unsettled, and this entry does not resolve it. The Sanskrit verse as printed names eight places for a group of seven [S2]. Four renderings were located. Wood's 1911 English gives Ayodhyā, Mathurā, Mayā, Kaśī, Kañchi, Avantikā, Dwārāvatī [S1]. Choudhury 1994 gives Ayodhya, Mathura, Kasi, Kanchi, Avantika, Puri, Dvaravati — keeping Puri, dropping Maya [S3]. Bhattacharyya 1987 gives an almost unrelated set, Conjeeveram, Dwaraka, Ayodhya, Mathura, Hardwar, Banaras, Ujjain, in which Dwarka is second, not seventh [S4]. A fourth set — Paraśurāma, Gokul, Haridwar, Prayag, Kashi, Ayodhya, Dwarka — circulates widely in devotional and tourism material; no source for it was located in this pass. The disagreement is recorded rather than adjudicated. Note that the corpus' own near-duplicate file dwarka_moksha_puri.md asserts that fourth set flatly as fact, with no source at all, which is the practice this entry is written against.
Most Tier-1 sources here were not read in the original. Google Books, archive.org full text, the Current Science archive and sacred-texts.com all refused automated access on 2026-09-29. Every affected source is marked NOT CONSULTED DIRECTLY in sources:. Only [S9], [S11] and [S12] — the two issuer pages, one of them the temple's own trust — were read in full. The claims the unconsulted sources support are reported as their claims, and the ones a reader might otherwise take as established — the 1473 sack, the 1559 image, the 43 m shikhara — should be checked against the printed works.
The temple trust's chronology and the archaeology disagree, and both are given. The trust places Vajranabh's monument in 400 BC and a Brahmi-script phase in 100 BC [S12]; the Archaeology Department's own opinion is 1200 years [S9]; the excavated material is early historical [S8][S15]. This entry does not arbitrate between a devotional-institutional document and an excavation report. Relatedly, the historical_period field records the 1473 sack, but Rangarajan's account of restoration implies a longer continuity of worship at the site than a single destruction allows [S10].
Removed, not quietly retained. The claim that underwater ruins had confirmed the submerged city, and that marine archaeology of the 1980s added "scientific legitimacy to the traditional belief", is withdrawn: the underwater work is real but does not confirm the city of the epic [S7][S8][S15]. "2+ million pilgrims annually" and a 30 km distance to Bet Dwarka were unsourced and are not repeated. The date of the city's submersion is given by no source reached. The claim that dying within a mokṣapurī's boundaries confers liberation, and that bathing at Gomati-tirtham confers it, are not asserted. Kapur's study treats the seven as a group whose pilgrimage confers mokṣa [S6], which is the general proposition and not the city-specific mechanism; the Sapta Puri article is listed as a finding aid only [S18] and supports no claim here.
This is a place, not a deity. The corpus files it under deities, and an earlier draft carried gender: female, which was a category error; it is now null. The deity, Dvārakādhīśa, has its own entry.
Sources
- [S1] Wood, E. and S. V. Subrahmanyam (transl.), Garuḍa Purāṇa, 1911, XVI:114 — the English rendering of the seven. Not consulted directly.
- [S2] Sahay, Shivswaroop, Pracheen Bharatiya Dharm Evam Darshan, Motilal Banarsidass, 2008, p. 151 — the Devanagari of the verse. Not consulted directly.
- [S3] Choudhury, Anuradha Roma, "Hinduism", in Sacred Place (eds. Holm & Bowker), Pinter, 1994, p. 70. Not consulted directly.
- [S4] Bhattacharyya, Sivaprasad, in K. W. Morgan (ed.), The Religion of the Hindus, Motilal Banarsidass, 1987, pp. 188–191. Not consulted directly.
- [S5] Bansal, Sunita Pant, Hindu Pilgrimage, Pustak Mahal, 2008, pp. 6–9, 34–35. Not consulted directly.
- [S6] Kapur, Anu, Mapping Place Names of India, Taylor & Francis, 2019, pp. 74–75. Not consulted directly.
- [S7] Gaur, A. S., Sundaresh and Sila Tripati, "An ancient harbour at Dwarka", Current Science 86(9), 2004. Not consulted directly — archive returned 404.
- [S8] Rao, S. R., Marine Archaeology of Indian Ocean Countries, NIO, 1988, pp. 18–25. Not consulted directly.
- [S9] Dwarkadhish Temple, Tourist Places, District Administration, Devbhumi Dwarka, Government of Gujarat — read in full, 2026-09-29.
- [S10] Rangarajan, Haripriya, Spread of Vaiṣṇavism in Gujarat up to 1600 CE, Somaiya, 1990, pp. 50–55, 51. Not consulted directly.
- [S11] Shree Dwarkadhish Temple official trust website, dwarkadhish.org — read in full, 2026-09-29 (chronology and published festival calendar).
- [S12] Dwarka Nagari — Little History, same trust site. Read in full, 2026-09-29.
- [S13] Dwarkadhish Temple, Wikipedia, oldid 1375996660 — finding aid only (SOURCES_POLICY rule 5), read to reach [S7]–[S16]; not cited as evidence.
- [S14] Ansari, Z. D. and M. S. Mate, Excavations at Dwarka: 1963, Deccan College, 1966, p. 23; with Chandra, Medieval India, 2004, p. 220. Not consulted directly.
- [S15] Tripathi, Alok, Remote Sensing and Archaeology, Sundeep Prakashan, 2005, p. 79. Not consulted directly.
- [S16] Harshananda, Swami, Hindu Pilgrim Centres, Ramakrishna Math, 2012, p. 87. Not consulted directly.
- [S17] The Mahabharata; and the Nalayira Divya Prabandham (āḻvār hymns) — cited as works, per charter, not at verse level.
- [S18] Sapta Puri, Wikipedia, oldid 1374169469 — finding aid only (SOURCES_POLICY rule 5), read to reach [S1]–[S6]; not cited as evidence.
