September 2026 · Monthly retrospective

A clearer record, not a simple weekly sum

The September edition brings late reporting, corrected records and approved additions into one calendar-month account, while keeping occurrence, faith identity, motive and counting status distinct.

1–30 September 2026 · Evidence cutoff 3 October · Reconciled 6 October.

The approved source set yields 43 retained baseline incidents, 42 incident units from the review queue and two additional Baha’i sentence-enforcement events. One linked-security incident is retained separately. Inclusion documents a relevant event or impact; it does not by itself prove religious motive.

Africa: civilian violence, with attribution kept visible

Nigeria accounts for 21 main incident records in this dataset. The record includes the overnight cathedral/community attack at Zaranga/Nzangula, the Ayilamo attack, later Plateau incidents and reported abductions. The Nzangula sources differ between 11 and 12 deaths, while Ayilamo’s eleven comprise seven civilians and four police; these figures must not be presented as individually verified Christian deaths. (persecution.org; pmnewsnigeria.com)

Two distinct reported attacks at Rubung and Federe share a combined two-death figure. Keeping the incidents separate does not permit assigning two deaths to each. The reported Borgu displacement and mosque impacts belong to a Muslim-context record, not an assumed anti-Christian campaign. (truthnigeria.com; truthnigeria.com)

The regional picture also includes the reported Kabu killing in Mozambique and the L’Ayaida worship raid in Algeria. Twelve people in the latter were detained or questioned, but that is not a claim that all twelve remained imprisoned or were convicted. (barnabasaid.org; meconcern.org)

Middle East: restriction and sentence enforcement

Iran’s four records encompass a reported Christian broadcasting restriction and three Baha’i sentence-enforcement incidents. The approved additions concern Iraj Shakour and Pedram Abhar entering prison on 7 September and Saba Sefidi on 10 September; they add two incidents and three custody people, not new fatalities. (e-n.org.uk; bic.org; hengaw.net; hengaw.net)

The Jerusalem school-permit case records September administrative impact rather than treating every affected employee or pupil as a separate incident. Hama’s cathedral explosion remains Reported, with device, attribution and motive qualifications preserved. (acninternational.org; opendoors.de)

Central Asia and the wider Asian record

The included source set does not supply a separate Central Asian incident; that is a limit of this record, not evidence that no persecution occurred there. China’s six records concern coercion, church operations and detention-related state action. Overlapping custody stages at Jiaquan and Jinzhai are not summed into a false unique-person total. (vomcanada.com; chinaaid.org)

India: violence and administrative pressure

India contributes 17 incidents, including worship disruption, threats, reported violence, legal action and the Pimpri-Chinchwad demolition. The two distinguishable 6 September West Bengal cases remain two incidents; the demolition remains one event despite reporting across weekly windows and conflicting accounts of judicial protection. (ucanews.com; ucanews.com; siasat.com)

A Hindu woman’s reported intimidation over praying to Jesus remains a religious-practice case without relabelling her Christian. Conversion allegations in custody cases are attributed allegations, not adjudicated facts. (altnews.in; aajtak.in)

Other regions: distinguish identity, motive and fire cause

Eleven Jewish-related incidents are included in September. Case-level evidence ranges from reported abuse and intimidation to police-supported investigations; later motive information in Belleville updates the 20 September attack rather than creating another incident. The wounded officer and attacker are not folded into a Jewish-congregant casualty claim. (abcnews.com)

The approved Placentia, Meadow Lake and Kilburn fires are counted, but suspicious or undetermined cause remains explicit. Two concern former churches. Their inclusion is not a determination of arson or religiously hostile targeting. (grc.ca; oca.org; hellorayo.co.uk)

Reading the impact figures

The main-incident ledger contains 62 deaths: 26 source-identified Christian, 24 in Christian contexts with individual faith unconfirmed, seven civilians of unestablished faith, four police and one Ahmadi victim. A further police death belongs to the separate linked-security incident. The cautious wounded subtotal is 19, with two additional reported beatings whose injury documentation is unclear.

The attributed abduction subtotal is 45, including an 18-person report with an official verification caveat. State custody or forced questioning totals 52 case-linked people; it is neither a current-prisoner census nor a claim that every case was a formal arrest. These metrics overlap and must not be added into a unique-person total. Each map record exposes the applicable source and numeric qualification.

What the map can and cannot show

Three entries use September discovery dates rather than known act dates. Exact locations are not verified attack-site coordinates: the map uses labelled city, district or regional locators, with country-level representation for the nationwide policy. A cluster number denotes incident records, not people killed or injured.

Earlier editions used full published weekly windows, while September uses the approved complete calendar-month source set. This edition therefore makes no direct monthly trend claim. Historical editions remain available as originally published, rather than being silently rewritten by this reconciliation.