Early Release – Genomic and Epidemiologic Insights into Ongoing Measles Outbreak, Israel, 2025–2026 – Volume 32, Number 9—September 2026 – Emerging Infectious Diseases journal

During April 7, 2025–February 28, 2026, a total of 3,203 measles cases, including 16 deaths, were reported across Israel, the largest outbreak since 2018–2019 (8). Epidemiologic investigations found 13 case-patients had a history of travel abroad shortly before illness onset (Table). Among reported case-patients, mean age was 5.58 years (SD 9.35 years; range 1 month–73 years); 82.7% were unvaccinated, 7.1% were partially vaccinated, 2.2% were fully vaccinated, and 8.5% had unknown vaccination status.

As part of routine surveillance, 132 measles virus–positive samples collected during April 2025–February 2026 were selected for sequencing (Appendix). Samples were sequenced on the basis of the following criteria: all available imported cases, all travel-associated cases, representative samples from newly affected areas, and randomized samples collected in established outbreak locations every 2 weeks throughout the outbreak. Samples were excluded from sequencing when epidemiologic data showed that transmission occurred in established measles-positive locations (i.e., same street or household) (Appendix). Complete measles virus genomes were sequenced (9) (Appendix) and used for phylogenetic analysis to support transmission chain reconstruction. The N450 region of the measles viral genome confirmed the sequences as genotype B3, matching the reference MVs/Quetta.PAK/44.20 strain from Pakistan.

Figure 1

Figure 1. Phylogenetic analysis from investigation of genomic and epidemiologic insights into ongoing measles outbreak, Israel, 2025–2026. Phylogenetic tree including 132 sequences from Israel collected during April 2025–February 2026 and 29 global…

Figure 2

Figure 2. Geographic distribution of sequenced measles cases and importation clusters from genomic and epidemiologic investigation into ongoing measles outbreak, Israel, 2025–2026. Map displays the geographic distribution of sequenced cases by city….

Phylogenetic analysis showed the Israel sequences clustered most closely with sequences from the Netherlands and the United States collected in 2025 (Figure 1); however, because up-to-date global sequence data in public repositories are limited, intermediate diversity could be missing. The Israel sequences showed 3 likely clusters within the dominant B3 lineage, each derived from a separate documented importation: Belgium, Vietnam, and Russia. We found no spread of D8 lineage sequences imported to Israel. The Russia cluster likely originated from a documented importation from Russia (patient 4; Rehovot, May 2025) and is linked to an additional sample collected in Haifa in June 2025. That cluster is most closely related to a global sequence from California, USA (2025), differing by 7 nucleotides. Transmission in the cluster was limited to only 2 cases because close contacts were fully vaccinated and that location has high overall vaccine coverage. The Vietnam cluster, separated from the nearest ancestor by 12 nt, likely originated from a documented importation from Vietnam (patient 3; Beit-Itzhak, April 2025) and is not linked to additional Israel cases in the phylogeny. No further transmission was documented in the cluster, consistent with sequencing results and similar to the Russia cluster. The main and largest cluster (n = 129) was traced through epidemiologic investigations to a documented importation from Belgium (patient 2; Modi’in Illit, April 2025). That case was diagnosed retrospectively through contact tracing of infected close family members. Although the imported case was confirmed serologically (positive IgM), the available specimen contained insufficient viral RNA, precluding sequencing. However, sequencing was available for an epidemiologically linked close contact of the imported case (Bnei-Brak, April 2025; Figure 1, red arrow), located within the main cluster; epidemiologic data indicated subsequent spread to additional cities in Central District, Jerusalem, and later North District. That cluster also includes an additional sequenced importation from Belgium (case 5; Jerusalem, June 2025), unrelated epidemiologically to the other Belgium cases, located on the ancestral branch. The main cluster is most closely related to global sequences from the Netherlands collected in 2025, differing by 2 mutations. That strain has subsequently spread in Israel and is still circulating, accumulating mutations and giving rise to several sublineages. Most samples in that cluster were from locations with low vaccine coverage and within Ultra-Orthodox communities (Figure 2; Appendix Table 3).

The sequences obtained from deceased children were located within different sublineages in the main outbreak cluster (Figure 1). The first case (Jerusalem, July 2025) was identical to 2 sequences detected elsewhere (Tel Aviv, September 2025; Emmanuel, August 2025). The second case (Jerusalem, July 2025) was identical to another from Jerusalem (June 2025). The third case (Poriya, November 2025) was identical to 2 cases from the North District (Nahariya, November 2025). The identical genomes suggest close epidemiologic linkage and, when detected in different districts, indicate transmission across multiple regions in Israel. That finding suggests the virus in fatal cases was typical of the outbreak strain. Three additional case-patients with documented travel abroad (patient 7, Poland; patient 10, Ethiopia; patient 12, Germany) near the time of measles detection had sequences within the main cluster, suggesting that infection was more likely acquired in Israel, not abroad.

Figure 3

Figure 3. Positivity rates by epidemiologic week during April 1–October 31, 2025, from genomic and epidemiologic investigation into ongoing measles outbreak, Israel, 2025–2026. Weekly number of measles cases by district was confirmed…

To investigate positivity dynamics from an epidemiologic perspective, we evaluated the weekly ratios of positive and negative samples collected during April–October 2025 (epidemiologic weeks 14–44). When the outbreak began, the positivity rate was low (10%–20%), likely reflecting heightened awareness and broad testing of suspected cases with low pretest probability for measles. As transmission became established in specific districts during weeks 28–38, positivity increased to ≈60%, indicating outbreak persistence and targeted sampling (Appendix). During weeks 39–44, as the outbreak expanded to additional districts, the overall ratio decreased (Figure 3); we observed a similar trend through February 2026 (data not shown).

The ongoing measles outbreak in Israel threatens the country’s elimination status, defined by the World Health Organization as absence of endemic transmission for >12 months (https://iris.who.int/server/api/core/bitstreams/b0ee4708-ba01-4b40-9003-f6f1aba317da/content). This outbreak highlights the profound risks associated with declining vaccination coverage. Genomic sequencing of targeted measles samples indicated that a single importation event was most likely the primary driver of the outbreak. Phylogenetic analysis showed 3 validated distinct importation events from Russia, Vietnam, and Belgium; outcomes differed drastically based on local immunity. The main Belgium-linked cluster strain, detected in April 2025, likely sparked an extensive national outbreak. The second documented importation case from Belgium might have contributed to the main transmission cluster. The epidemiologic investigation indicated that this case-patient had many contacts and resided in an area experiencing a substantial outbreak. However, the case was detected 2 months after the Belgium index case, so it likely had little impact on outbreak dynamics. Furthermore, its shared mutations with the April 2025 Belgium case suggest a common source.

Additional undetected single cases might exist because of underreporting. However, our strict criteria and direct contact to Ministry of Health epidemiologic data enabled systematic assessment of each focal point, so underreporting is unlikely to have substantially affected the main phylogenetic findings. Positivity rates rose from 10%–20% during early broad testing to 60% during weeks 28–38, as diagnostics shifted to targeted surveillance during outbreak spread.

In summary, we integrated epidemiologic investigations with molecular surveillance, demonstrating that a single case can spark extensive measles transmission. Our findings highlight the critical role of molecular surveillance in measles outbreak management.