Attempts to date the events using methods of archaeoastronomy have produced, depending on which passages are chosen and how they are interpreted, estimates ranging from the late 4th to the mid-2nd millennium BCE.[52] The late 4th-millennium date has a precedent in the calculation of the Kali Yuga epoch, based on planetary conjunctions, by Aryabhata (6th century). Aryabhata's date of 18 February 3102 BCE for Mahābhārata war has become widespread in Indian tradition. Some sources mark this as the disappearance of Krishna from the Earth.[53] The Aihole inscription of Pulakeshin II, dated to Saka 556 = 634 CE, claims that 3,735 years have elapsed since the Bhārata battle, putting the date of Mahābhārata war at 3137BCE.[54][55]
Another traditional school of astronomers and historians, represented by Vrddha Garga, Varāhamihira and Kalhana, place the Bharata war 653 years after the Kali Yuga epoch, corresponding to 2449 BCE.[56] According to Varāhamihira's Bṛhat Saṃhitā (6th century), Yudhishthara lived 2,526 years before the beginning of the Shaka era, which begins in the 78 CE. This places Yudhishthara (and therefore, the Mahabharata war) around 2448–2449 BCE (2526–78). Some scholars have attempted to identify the "Shaka" calendar era mentioned by Varāhamihira with other eras, but such identifications place Varāhamihira in the first century BCE, which is impossible as he refers to the 5th century astronomer Aryabhata. Kalhana's Rajatarangini (11th century), apparently relying on Varāhamihira, also states that the Pandavas flourished 653 years after the beginning of the Kali Yuga; Kalhana adds that people who believe that the Bharata war was fought at the end of the Dvapara Yuga are foolish.[57]