The ESV 2025: Its Stated Hebrew and Greek Bases

Revision note: Substantially revised July 21, 2026. Edition and publication claims were checked through July 17, 2026; the argument was compressed and its research apparatus consolidated below.

The English Witness Must Be Identified Exactly

This series uses the English Standard Version to let readers see the practical English form of many decisions associated with modern critical editions. That convenience creates a danger. “The ESV” can become a vague label for several different things:

• an English translation first published in 2001;
• a revision lineage with updates in 2007, 2011, 2016, and 2025;
• a translation whose wording begins from the Revised Standard Version tradition;
• a committee’s interpretation of Hebrew, Aramaic, and Greek;
• a body text accompanied by textual footnotes;
• hundreds of physical editions moving through a multi-year production cycle; or
• a supposed English transcription of one fixed “critical text.”

Only the first six descriptions can be defended. The seventh cannot.

The controlling English witness for this project is the ESV 2025 Text Edition, including its textual footnotes wherever available. Crossway’s current preface names the source editions on which it is based. That statement is sufficiently precise to establish a formal baseline, and sufficiently qualified to prevent us from treating the translation as a mechanical reproduction of that baseline.

What Crossway Officially Names

Under “Textual Basis and Resources,” the ESV preface identifies three printed source editions:

Crossway also says that the Old Testament committee sought, where possible, to translate difficult Masoretic readings as they stand rather than quickly emending them. In exceptional difficult cases it consulted the Dead Sea Scrolls, Septuagint, Samaritan Pentateuch, Syriac Peshitta, Latin Vulgate, and other sources. The preface permits those sources to illuminate the Masoretic wording or, when judged necessary, to support a departure from it.¹

The New Testament statement contains a parallel qualification. The ESV is based upon UBS5/NA28, but in a few difficult cases it follows a Greek reading other than the one preferred in those editions.²

Those two admissions are not incidental disclaimers. They are part of the textual basis.

“Based On” Is a Relationship, Not an Equation

If (E) represents the English wording and (S) the stated source edition, the preface asserts a general relationship (E \leftarrow S). It does not assert the identity (E = S).

An English translation cannot literally equal a Hebrew or Greek edition. It changes language, and therefore must make decisions concerning lexical range, syntax, discourse, idiom, ambiguity, punctuation, and style. More importantly for this study, the committee acknowledges that it sometimes selects a textual reading other than the one printed in the source edition’s main line.

The phrase based on therefore has at least four components:

1. default corpus: the edition ordinarily consulted as the running source;
2. textual exceptions: places where another reading is followed;
3. interpretive decisions: places where the source wording permits several English renderings;
4. translation inheritance: wording carried forward, revised, or reconsidered from the RSV and the older Tyndale–King James stream.

The first component establishes the presumption. The other three prevent the presumption from becoming an untested conclusion at an individual verse.

The Old Testament Baseline: BHS 1997

BHS remains the last complete scholarly edition in the Biblia Hebraica series. Its running text is based on Codex Leningradensis, while its apparatus records variants and editorial proposals. The German Bible Society explicitly contrasts this design with the reconstructive aim of scholarly Greek New Testaments.³

The ESV’s use of BHS means that the Masoretic text of Leningrad normally stands behind its Old Testament. It does not mean that every English difference from the King James Version reflects a different Hebrew consonantal text. Many differences arise from:

• a different lexical judgment;
• a different analysis of syntax;
• a different treatment of a Hebrew idiom;
• a different vocalization or word division;
• a decision about singular and collective reference;
• English modernization;
• a change in paragraphing or punctuation; or
• a different translation philosophy.

Conversely, some differences do represent a textual departure. BHS itself may cite the evidence, but its running line generally remains Leningrad. The ESV can adopt an alternative in English while the BHS main text above the apparatus remains unchanged.

Every Old Testament dossier must therefore ask two separate questions: “What Hebrew does BHS print?” and “What textual or translational decision produced the ESV wording?”

The ESV preface’s respect for difficult Masoretic readings is methodologically important. It gives the Masoretic main text a stated presumption. It is not an absolute prohibition on emendation, nor does it identify which exceptions occur. The footnotes help locate some of them, but the series must verify each claim from the Hebrew and the edition apparatus rather than infer the answer from the English alone.

The New Testament Baseline: UBS5 and NA28

UBS5 and NA28 were designed for different primary audiences and carry different apparatuses, but their Greek main text belongs to the same editorial text at the relevant edition state. NA28 incorporates the ECM-based revision of the Catholic Epistles; UBS5 incorporates that same text while presenting a translator-oriented selection of variants.

Crossway’s preface names both because a translation committee needs both kinds of resource. NA28 supplies a compact scholarly apparatus across the New Testament. UBS supplies a more selective apparatus focused on variants judged significant for translation, with its own editorial aids. Naming both does not mean that the ESV combined two competing continuous Greek texts verse by verse.

It also does not make UBS5 or NA28 current in every sense in 2026. UBS6 has now published ECM-derived changes in Mark, Acts, and Revelation, and it functions as the available main-text proxy for portions destined for NA29.⁴ NA29 itself remains in preparation. Yet Crossway’s current ESV preface still names UBS5 and NA28. The ESV 2025 did not silently become a translation of UBS6 merely because UBS6 appeared.

That chronological separation must be preserved:

• current ESV stated base: UBS5 corrected and NA28;
• current published scholarly Greek control in this series: NA28 plus UBS6 where its main text advances beyond NA28;
• future edition: NA29, when published and examined;
• translation update: ESV 2025, which does not automatically incorporate every later Greek editorial decision.

A future ESV update could revise the stated base. Until Crossway says so, this project records what the current preface actually says.

What Changed in the 2025 Text Edition

Crossway announced the 2025 revision on February 11, 2025. Its official summary reports body-text changes in thirty-six passages involving forty-two verses and sixty-eight words. It also reports changes to fifty-seven footnotes and punctuation changes in fourteen verses.⁵

The scale is deliberately limited. The committee says it values a stable standard text and does not plan frequent updates. It also remains open to revisions prompted by archaeological, linguistic, and textual evidence.

Two public examples show why a translation edition cannot be reduced to a source-text edition.

In Genesis 3:16 and 4:7, the 2025 ESV returned from “contrary to” to “for” in rendering the Hebrew preposition ʾel, changed “but” to “and,” and expanded the interpretive options in the footnotes. The underlying BHS consonantal text did not change. The English interpretation did.⁶

In John 1:18, the ESV changed “the only God” to “God the only Son” and revised the textual footnote. This is a consequential Christological rendering, yet Crossway presents it as an English decision involving theos, monogenēs, context, concordance, and manuscript alternatives—not as the adoption of a newly published Greek base edition.⁷

The complete official comparison document identifies body, footnote, and punctuation revisions side by side.⁸ It must be consulted whenever a later article depends upon a wording changed in 2025. A quotation from an older ESV cannot be silently treated as the 2025 witness.

Translation Edition, Source Edition, and Product Rollout

Three timelines are easy to confuse.

The source-edition timeline

• NA28: 2012.
• UBS5: 2014, with corrected state named by Crossway.
• BHS fifth corrected edition: 1997.
• UBS6: subsequent current Greek edition, not named as the ESV 2025 base.
• BHQ: an ongoing series of fascicles, not named as the ESV base.

The translation-edition timeline

• ESV first edition: 2001.
• intervening revisions: 2007, 2011, and 2016.
• current controlling text edition: 2025.

The manufacturing and distribution timeline

Crossway stated that the 2025 text would take nearly two years to roll through more than 540 editions, from spring 2025 toward fall 2026.⁹ On the July 17, 2026 cutoff, a newly purchased physical ESV could still require inspection of its copyright page. “Printed in 2026” does not itself prove “ESV Text Edition: 2025.” Inventory, typesetting, and format transitions do not occur simultaneously.

For reproducibility, this series will rely upon a witness explicitly marked ESV Text Edition: 2025 or the current official digital text, and it will record the source when wording matters.

The Footnotes Are Part of the Evidence

Crossway calls its textual footnotes an integral part of the translation. They alert readers to variants and difficulties, disclose some decisions, and offer significant alternative readings.¹⁰

That statement has two implications.

First, an audit that quotes only the ESV body can omit the committee’s own qualification. A body reading may look exclusive while the footnote acknowledges another textual form. For a reader-facing analysis, body and footnote belong together.

Second, a footnote does not place every alternative on the same level. It may report another manuscript reading, a literal rendering, an alternate word division, a versional reading, an interpretive possibility, or an explanatory gloss. Later articles must classify the note rather than simply count it as “another Bible.”

The dataset will therefore transcribe the ESV body and footnote into separate fields and assign each footnote a type.

What the ESV Can and Cannot Establish

The ESV can establish what a prominent modern essentially literal translation presents to an English reader. It can reveal where the translation committee judged a textual difference important enough to affect the body or a note. It can show how a critical-edition decision functions in an ordinary English sentence.

The ESV cannot by itself establish:

• the precise Greek or Hebrew reading adopted;
• whether the English difference is textual or translational;
• whether an alternative appears in the source edition’s main text or only its apparatus;
• the strength and distribution of manuscript support;
• whether a rendering is required by the source wording;
• whether the same decision appears in UBS6 or BHQ; or
• whether a Westminster proposition rises or falls with the verse.

Those claims require the original-language controls and the confessional proof ledger.

The Strongest Defense of Using the ESV as “the Critical Text in English”

There is a legitimate shorthand behind the phrase. The ESV is an essentially literal translation; it names BHS, UBS5, and NA28 as its base; it aims for transparency; and its committee worked directly with the original languages. For many common readings, it gives a fair and readable representation of the wording adopted in those scholarly editions.

This series will use that practical advantage. It will not use the shorthand as an identity claim.

At disputed places, the ESV is best called the reader-facing English witness to a critical-edition-based translation process. That longer description is less elegant and more accurate. It leaves room for precisely the exceptions Crossway acknowledges.

Confessional Use of an English Translation

Westminster Confession 1.8 requires Scripture to be translated into the ordinary language of every nation. It also locates immediate inspiration and final authenticity in the Hebrew Old Testament and Greek New Testament, not in one later vernacular rendering.¹¹

Using the ESV is therefore not a concession that English governs the originals. It is an act of communication under an original-language control. When the ESV faithfully renders the selected Hebrew or Greek, it helps the reader follow the audit. When it obscures or departs from that source, the audit must say so.

This also protects the series from a false symmetry. Confessional Bibliology does not require treating the King James Version as an inspired correction of Hebrew and Greek, and this project will not grant that role to the ESV. Both are translations. Their textual bases, translation decisions, and historical reception differ, but neither becomes the autographic language of Scripture.

ESV Control Rules Adopted for the Series

The following rules now govern every later use of the ESV:

1. “ESV” means the 2025 Text Edition unless another edition is expressly named.
2. Body text and textual footnotes are recorded separately.
3. The presumed Old Testament base is BHS 1997, subject to verification at each disputed place.
4. The presumed New Testament base is UBS5/NA28, subject to verification at each disputed place.
5. Agreement in English does not prove identity in the source text.
6. Difference in English does not prove difference in the source text.
7. Every suspected departure receives a classification: direct witness, versional reconstruction, alternate vocalization, conjecture, translation choice, or unresolved.
8. UBS6 and BHQ are examined independently; their later publication does not silently update the ESV.
9. A change between ESV 2016 and 2025 is logged as translation-edition drift even if the source base remained the same.
10. Doctrinal effects are scored only after the source and translation layers are separated.

Verdict

It is established that the ESV 2025 officially names BHS 1997 as its Old Testament base and UBS5 corrected/NA28 as its New Testament bases. It is equally established, from the same preface, that the ESV sometimes consults and follows evidence outside those running texts.

The ESV 2025 is therefore adopted as the project’s reader-facing English witness, not as a transparent substitute for the Hebrew and Greek editions. Its textual footnotes are included, its edition state is fixed, and every consequential wording will be checked against the named original-language controls.

No A–F doctrinal grade is assigned because this article establishes the English witness rather than evaluating a Westminster proposition.

What This Does Not Prove

The ESV’s use of modern critical editions does not prove that every ESV reading is critically correct, confessionally defective, or doctrinally neutral. Its selective departures do not prove arbitrariness. Its 2025 revisions do not prove that the translation is unstable in ordinary use.

The argument is narrower: a stated base establishes a rebuttable textual presumption, not a verse-by-verse equation between an English translation and a Greek or Hebrew running text.

Text and Edition Record

• English witness: ESV, 2025 Text Edition.
• Old Testament stated base: Masoretic Text in BHS, fifth corrected edition, 1997.
• New Testament stated bases: UBS Greek New Testament, fifth corrected edition (2014), and NA28 (2012).
• Acknowledged Old Testament supplementary witnesses: Dead Sea Scrolls, Septuagint, Samaritan Pentateuch, Peshitta, Vulgate, and other sources.
• Acknowledged New Testament exception: A few places follow a Greek reading not preferred by UBS/NA28.
• 2025 update scope reported by Crossway: Thirty-six passages, forty-two verses, sixty-eight body-text word changes; fifty-seven footnotes; punctuation in fourteen verses.
• Current later Greek control: UBS6, examined independently rather than projected backward into the ESV.
• Current Hebrew successor: BHQ fascicles, examined independently rather than treated as the ESV base.
• Status cutoff: July 17, 2026.

Notes and Further Reading

• Crossway. “ESV Bible Translation Update.” February 11, 2025.
• Crossway. ESV Text Edition: 2025—Complete List of Changes from 2016. 2025.
• Crossway. “Preface to the English Standard Version.”
• Deutsche Bibelgesellschaft. “Biblia Hebraica Stuttgartensia (BHS).”
• Deutsche Bibelgesellschaft. “The 28th Edition of the Novum Testamentum Graece.”
• Deutsche Bibelgesellschaft. “The Sixth Edition of the Greek New Testament (UBS6).”
• Westminster Assembly. The Westminster Confession of Faith. 1647.

Revision Record

• Version 1.0 — July 17, 2026: Fixed the ESV 2025 as the English witness, documented its stated bases and admitted exceptions, and adopted translation-edition controls.