I was trying to answer Which scripture was first to identify Buddha as an avatar of Lord Vishnu? (Hinduism SE) but then I don't know which is the oldest of the Purāṇas.

I'm not looking for references from Purāṇas that call either themselves or others the oldest but research done on this subject from a historical standpoint, i.e., who wrote these Purāṇas and when (on human timescale), the kind of Sanskrit script that was used for writing the oldest parts of the Purāṇas, etc.

Questions that are related but don't address my question:

PS. The moderators at Hinduism SE deleted my own answer (posted below) to this question. They also denied my request to migrate both the Q & A to History SE, so I'm doing this myself.

  • 2
    How is this historical rather than theological? Historiography is a limited methodology, whose claims regarding supposedly sacred texts are normal rejected for lack of sacrality. Commented May 9, 2022 at 11:00
  • 2
    Not theological because Hindus generally believe events described in all Puranas keep repeating over the Hindu time span and all 18 of them were composed by the same person Vyasa. So asking which among them was the first from theological perspective is kind of meaningless/controversial. @SamuelRussell
    – sv_
    Commented May 9, 2022 at 16:23
  • 2
    Historians and secular scholars on the other hand don't believe this. They think the 18 Puranas were written by different authors at different times. That's why I think this Q belongs here and safe from censorship. @SamuelRussell
    – sv_
    Commented May 9, 2022 at 16:24

1 Answer 1


According to the list of Purāṇas provided in A Concise Encyclopaedia Of Hinduism by Swami Harshananda, Vāyu Purāṇa from A.D. 200 appears to be the oldest.

| S. No.  | Name of the     | Number of  | Period of     |
|         | Purāṇa          | Ślokas     | Composition   |
|         |                 |            |               |
|    1.   | Agni            |    16,000  | A.D. 800      |
|    2.   | Bhāgavata       |    18,000  | A.D. 600      |
|    3.   | Bhaviṣya        |    14,500  | A.D. 500-900  |
|    4.   | Brahma          |    10,000  | A.D. 1300     |
|    5.   | Brahmāṇḍa       |    12,000  | A.D. 400      |
|    6.   | Brahmavaivarta  |    18,000  | A.D. 1000     |
|    7.   | Garuḍa          |    18,000  | A.D. 900      |
|    8.   | Kūrma           |    18,000  | A.D. 500      |
|    9.   | Liṅga           |    11,000  | A.D. 600-1000 |
|   10.   | Mārkaṇḍeya      |     9,000  | A.D. 300      |
|   11.   | Matsya          |    14,000  | A.D. 300      |
|   12.   | Nāradīya        |    25,000  | A.D. 900-1600 |
|   13.   | Padma           |    55,000  | A.D. 800      |
|   14.   | Skanda          |    81,000  | A.D. 700-900  |
|   15.   | Vāmana          |    10,000  | A.D. 900      |
|   16.   | Varāha          |    24,000  | A.D. 800-1100 |
|   17.   | Vāyu            |    24,000  | A.D. 200      |
|   18.   | Viṣṇu           |    23,000  | A.D. 300      |

The author however notes "the period of compilation is very approximate."

P. L. Bhargava in his paper The Origin and Development of Purāṇas and Their Relation With Vedic Literature lists Vāyu, Brahmāṇḍa, Matsya, Viṣṇu and Mārkaṇḍeya Purāṇas among the oldest but doesn't say which of them is the oldest.

The internal evidence of the Purāṇas makes it absolutely clear that they were written at widely different times. As a rule, the later the Purāṇa, the more its historical matter is burdened with explanatory myths and the latest Purāṇas are altogether relieved of historical matter. Reversely, the older the Purāṇa, the less its genealogies are burdened with myths and the more they approximate to the Vedic evidence. Judged from this, the oldest Purāṇas, so far as genealogies and accounts of persons mentioned in the genealogies are concerned, are the Brahma, the Vāyu, the Brahmāṇḍa and the Matsya, all of which purport to be narrated by the sūta Lomaharṣaṇa. Later than these four but older than the others is the Viṣṇu Purāṇa. Seven other Purāṇas, the Bhāgavata, the Garuda, the Agni, the Padma, the Liṅga, the Kūrma and the Mārkaṇḍeya contain historical matter in varying degree. Among the remaining Purāṇas, the Bhaviṣya professedly does not deal with the ancient past and the other five viz. the Nārada, the Brahmavaivarta, the Varāha, the Vāmana and the Skanda are purely sectarian works free from all historical matter.

The chronology of the Purāṇas, however, is not as easy as this. For instance the Brahma, whose genealogical account is ancient, contains other material which is comparatively quite late. On the other hand the Mārkaṇḍeya whose genealogical material is not of much value yet appears to be quite ancient if we judge it from the other material that it contains. It is the least sectarian of the Purāṇas which tended to become more and more sectarian with the passage of time. If we cast a glance over the entire contents of the Purāṇas, the oldest Purāṇas appear to be the Vāyu, the Brahmāṇḍa, the Matsya, the Viṣṇu and the Mārkaṇḍeya all of which must have been composed by the 4th century A. D. although the Matsya and the Mārkaṇḍeya contain some material which appears to be much later. The first four of these Purāṇas contain the lists of kings who ruled in the Kali age and this list ends with the foundation of the Gupta dynasty in the beginning of the 4th century A. D.

There is another way of judging the relative chronology of some of the Purāṇas. The Viṣṇu contains the oldest version of the Kṛṣṇa story. The versions of the Brahma and the Bhāgavata are later and those of the Padma and the Brahmavaivarta which mention the name of Rādhā are later still. This proves that the Brahma to a large extent and the Bhāgavata wholly are later than the Viṣṇu, while the Padma and the Brahmavaivarta came later still. Since Alberuni mentions all the Purāṇas they must all have been composed by 1000 A. D.

  • 1
    Theological, not historiographical. Commented May 9, 2022 at 11:02
  • Compilation, or composition? In your table header you have "Period of composition" but the quote from the author just below it mentions "compilation." Commented May 9, 2022 at 21:29
  • 2
    Must be a typo. I think he means composition. Because he uses 'compilation' (for the original author) and 'composition' (for his disciples' works) at this link: "The sage Vedavyāsa might have compiled these into one Purāṇasaṁhitā. His disciples and their disciples as also others in that tradition might have composed more detailed works which gradually took the present form, the eighteen purāṇas as we know them today." @DavidConrad
    – sv_
    Commented May 9, 2022 at 21:38

Not the answer you're looking for? Browse other questions tagged or ask your own question.