diff --git a/pipeline/plugins/extractors/tap-uk-ees/tap_uk_ees/tap.py b/pipeline/plugins/extractors/tap-uk-ees/tap_uk_ees/tap.py
index d0d9300..b7557e4 100644
--- a/pipeline/plugins/extractors/tap-uk-ees/tap_uk_ees/tap.py
+++ b/pipeline/plugins/extractors/tap-uk-ees/tap_uk_ees/tap.py
@@ -40,17 +40,36 @@ def _slug_to_time_period(slug: str) -> str | None:
 
 
 def get_all_releases(publication_slug: str) -> list[dict]:
-    """Return all releases for a publication as dicts with 'id' and 'time_period'."""
-    url = f"{CONTENT_API_BASE}/publications/{publication_slug}/releases"
-    resp = requests.get(url, timeout=TIMEOUT)
-    resp.raise_for_status()
-    data = resp.json()
-    # API returns either a plain list or a paginated object with a "results" key
-    releases = data if isinstance(data, list) else data.get("results", [])
+    """Return all releases for a publication as dicts with 'id' and 'time_period'.
+
+    The EES content API paginates with a 'paging' envelope when there are many
+    releases. This function follows all pages so no historical release is missed.
+    """
     result = []
-    for r in releases:
-        time_period = _slug_to_time_period(r.get("slug", ""))
-        result.append({"id": r["id"], "time_period": time_period})
+    page = 1
+    while True:
+        url = f"{CONTENT_API_BASE}/publications/{publication_slug}/releases?page={page}&pageSize=20"
+        resp = requests.get(url, timeout=TIMEOUT)
+        resp.raise_for_status()
+        data = resp.json()
+
+        # API returns either a plain list or a paginated object with a "results" key
+        if isinstance(data, list):
+            releases = data
+            total_pages = 1
+        else:
+            releases = data.get("results", [])
+            paging = data.get("paging", {})
+            total_pages = paging.get("totalPages", 1)
+
+        for r in releases:
+            time_period = _slug_to_time_period(r.get("slug", ""))
+            result.append({"id": r["id"], "time_period": time_period})
+
+        if page >= total_pages:
+            break
+        page += 1
+
     return result
 
 
diff --git a/pipeline/transform/models/staging/stg_ees_ks4.sql b/pipeline/transform/models/staging/stg_ees_ks4.sql
index a99c8d9..7cd1d3e 100644
--- a/pipeline/transform/models/staging/stg_ees_ks4.sql
+++ b/pipeline/transform/models/staging/stg_ees_ks4.sql
@@ -3,9 +3,11 @@
 -- Staging model: KS4 attainment data from EES
 -- KS4 performance data is long-format with breakdown dimensions (breakdown_topic,
 -- breakdown, sex). Unlike KS2 which has a subject dimension, KS4 metrics are
--- already in separate columns — we just filter to the 'All pupils' breakdown.
+-- already in separate columns — we just filter to the all-pupils total row.
 -- EES uses 'z' (not applicable) and 'c' (confidential) as suppression codes —
 -- safe_numeric handles both by treating any non-numeric string as NULL.
+-- NOTE: older EES releases (pre-2023/24) use breakdown_topic = 'All pupils';
+-- the 2023/24 release switched to breakdown_topic = 'Total'. Both are included.
 
 with performance as (
     select * from {{ source('raw', 'ees_ks4_performance') }}
@@ -46,7 +48,7 @@ all_pupils as (
         {{ safe_numeric('gcse_91_percent') }} as gcse_grade_91_pct
 
     from performance
-    where breakdown_topic = 'Total'
+    where breakdown_topic in ('Total', 'All pupils')
       and breakdown = 'Total'
       and sex = 'Total'
 ),