Take heed therefore unto yourselves, and to all the flock, over the which the Holy Ghost hath made you overseers, to feed the church of God, which he hath purchased with his own blood.
Why the KJV Reading Is Retained
The KJV’s “church of God” is supported by Sinaiticus, Vaticanus, and a substantial Greek manuscript group. On this principal noun, the KJV agrees with the modern critical text against both the “Lord” reading and the later Byzantine conflation “Lord and God.”
Disputed unit
Two independent units: θεοῦ/κυρίου and conflations at the church title, followed by ἰδίου αἵματος versus αἵματος τοῦ ἰδίου.
The selected ECM corpus is multiway at the title: 34 Greek states read God, 34 Lord, 21 Lord and God plus Byz, and two singleton Jesus-inclusive forms. At the separate blood-order unit, 29 named entries have the KJV order plus Byz and 49 have the competing order.
Primary support for the KJV reading
Ancient Greek, Syriac, Coptic, Latin and patristic evidence for God, with Byzantine support for the KJV word order
“
King James Bible reading
Acts 20:28 in the King James Bible
Take heed therefore unto yourselves, and to all the flock, over the which the Holy Ghost hath made you overseers, to feed the church of God, which he hath purchased with his own blood.
Early second century — Ignatius, related phrase “blood of God.” This is theological and verbal support, not an Acts quotation. Exact Greek “Church of God”: fourth century — Codices Sinaiticus and Vaticanus.
Earliest surviving Greek support: Fourth century — Codices Sinaiticus and Vaticanus.
Primary support for the KJV reading
Ancient Greek, Syriac, Coptic, Latin and patristic evidence for God, with Byzantine support for the KJV word order
The sections below preserve each supporting, competing, and related record under its proper evidence type.
Majority Text standing
The order “his own blood”
At the blood-order unit, the KJV/TR order is the Byzantine majority reading.
Estimated Greek support: Approximately 450 to 550 Greek manuscripts in the wider Acts tradition
Unit qualification: This majority claim applies only to the order of “his own blood.” At the divine-title unit, the Byzantine reading is the conflated “Lord and God,” not the exact KJV “God.”
This is a conservative working estimate for the wider Acts manuscript tradition. The exact source-scoped count remains printed beside the roster, and lacunose copies are not treated as votes.
Source scope and denominator
The current digital ECM Acts apparatus. Counts describe the named reading states displayed in its selected corpus; Byz remains an unexpanded group.
Greek support in the cited corpus
34 God states; 29 KJV-order states plus Byz
Greek competing evidence in the cited corpus
34 Lord states; 49 competing-order entries
Principal competing witnesses named in the snapshot
P41
P74
Codex Alexandrinus (02)
Codex Bezae (05)
Evidence represented on this page
Greek manuscripts
Represented
Ancient versions
Represented
Church fathers
Represented
Printed editions
Represented
Complete source-scoped roster
Complete roster within the stated source scope.
This roster preserves every named or grouped state recovered in the governing sources, including exact readings, competing readings, corrections, related forms, lacunae, versions, and aggregate traditions. When Maj or Byz appears on the KJV side, it marks the numerically dominant Greek transmission at that unit. The green Majority Text card above supplies the best available passage-specific estimate. The grouped siglum is then retained so it is not mistaken for one individual manuscript.
Apparatus and roster scope
The current digital ECM Acts apparatus. Counts describe the named reading states displayed in its selected corpus; Byz remains an unexpanded group.
Greek support in cited scope
34 God states; 29 KJV-order states plus Byz
Greek competing evidence in cited scope
34 Lord states; 49 competing-order entries
Evidence classification
Ancient Greek, Syriac, Coptic, Latin and patristic evidence for God, with Byzantine support for the KJV word order
Precisely located citations and additional names reported by the controlling apparatus are kept separate. An apparatus-level attribution is not presented as an independently verified exact quotation.
Ignatius of Antioch
AD 101–133
Letter to the Ephesians 1
Supporting witnessCitation type: Literary or theological use
uses “blood of God,” showing early Christian use of that expression but not quoting Acts 20:28
The noun and word-order units are independent and must not be combined into one vote.
424* has no assigned noun reading; [424C] reads Lord and God.
Evidence notes
Important distinctions in the evidence.
The Byzantine majority’s “Lord and God” is a conflated third reading, not the exact KJV form.
The construction “his own” may be interpreted as referring to God’s own Son, but that interpretation is not a separate manuscript reading.
Ignatius provides a theological parallel rather than direct evidence for the wording of a copy of Acts.
The noun and word-order units are independent and must not be combined into one vote.
424* has no assigned noun reading; [424C] reads Lord and God.
Evidence assessment
Why the KJV reading is retained
The selected ECM corpus is multiway at the title: 34 Greek states read God, 34 Lord, 21 Lord and God plus Byz, and two singleton Jesus-inclusive forms. At the separate blood-order unit, 29 named entries have the KJV order plus Byz and 49 have the competing order.
Primary support for the KJV reading
Ancient Greek, Syriac, Coptic, Latin and patristic evidence for God, with Byzantine support for the KJV word order
Evidence represented on this page
Evidence types, without turning rows into votes.
GreekRepresented
LatinRepresented
Other versionsRepresented
FathersRepresented
Printed editionsRepresented
Separate evidence category
Printed editions and reception history
Printed editions are presented apart from handwritten manuscripts and ancient versions. Edition-group summaries appear first; named supporting and competing editions then follow chronologically.
Edition evidence for both readings
Sources for the dates and readings in this section