Before and around the beginning of the Common Era: The work drew together an older medical inheritance and a wider field of ascetic and philosophical argument. Its Ātreya–Agniveśa teaching frame locates knowledge in remembered instruction and scholarly dialogue. It cannot by itself prove that the named assemblies occurred as narrated.
About 100 BCE–150/200 CE: This is the defensible scholarly range for the redactor called Caraka, not for every sentence in the received compendium. Different vocabulary, arguments, and borrowed materials indicate that the work contains more than one chronological stratum (Wujastyk 2012; Vienna Caraka Project).
Later completion: The received text credits Dṛḍhabala with supplying 17 of the 30 chapters in Cikitsāsthāna and all 12 chapters each of Kalpasthāna and Siddhisthāna—41 chapters in total. This internal attribution identifies a major layer; it does not securely date Dṛḍhabala or reveal how closely the supplied chapters resembled lost material (NIIMH e-Samhita).
By the eleventh century: Cakrapāṇidatta's Āyurvedadīpikā (“Lamp on Ayurveda”) explained terms, chose readings, and related passages to a wider scholastic tradition. It is the oldest completely preserved commentary on the compendium. Many modern editions print it with the root text, so its interpretations can quietly shape what readers take “Caraka” to mean (Vienna Caraka Project).
Manuscript centuries: Copyists, patrons, physician-scholars, families, and seats of learning reproduced the work in different scripts and regions. Cristina Pecchia's study treats these manuscripts as products of particular intellectual and social settings, not transparent containers for a single unchanging original. Manuscript comparison reveals variant readings and cross-contamination—copies made from more than one exemplar (Pecchia 2009; Vienna Caraka Project).
1868 and after: Gangadhar Ray Kaviraj began an edition with his Jalpakalpataru commentary at the Samvadajnanaratnakar Press in Kolkata. It appeared in Bengali script and was duplicated in Devanagari; later fascicles and editions continued the undertaking. Print widened and regularised access, but publishers, editors, script, and chosen manuscripts still determined the text readers received (Pecchia 2022).
1941 to the present: Jadavji Trikamji Acharya's Bombay edition with Cakrapāṇidatta became a widely used printed text, but a standard or “vulgate” edition is not the same thing as a critical edition reconstructed from all relevant witnesses. The University of Vienna project has assembled a major digital manuscript archive and applies stemmatic analysis to selected chapters, demonstrating how much textual work remains behind any confident translation (Wujastyk 2012; Vienna Caraka Project).