diff --git a/chromium_src/components/browsing_data/content/browsing_data_test_util.cc b/chromium_src/components/browsing_data/content/browsing_data_test_util.cc
new file mode 100644
index 00000000000..5f9d658bfdd
--- /dev/null
+++ b/chromium_src/components/browsing_data/content/browsing_data_test_util.cc
@@ -0,0 +1,40 @@
+/* Copyright (c) 2026 The Brave Authors. All rights reserved.
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this file,
+ * You can obtain one at https://mozilla.org/MPL/2.0/. */
+
+// Brave disables Chromium's cookie IPC cache optimization (because Ephemeral
+// Storage can switch the cookie backend at runtime), which means
+// document.cookie always uses IPC.  When a test sets a cookie via
+// document.cookie and immediately queries the backend cookie store through a
+// *different* Mojo pipe (CookieManager::GetAllCookies), there is no ordering
+// guarantee between the two pipes.  Upstream Chromium sees this as a ~0.6 %
+// flake; Brave amplifies it because every cookie read/write goes through IPC.
+//
+// Fix: after ExecJs("set<Type>()"), verify the data is accessible via
+// EvalJs("has<Type>()").  For cookies this forces a round-trip through
+// RestrictedCookieManager → CookieStore, serialising with the prior
+// SetCanonicalCookieAsync on the same CookieMonster task runner and
+// guaranteeing the cookie is committed before any subsequent GetAllCookies.
+
+#define SetDataForType SetDataForType_ChromiumImpl
+
+#include <components/browsing_data/content/browsing_data_test_util.cc>  // IWYU pragma: export
+
+#undef SetDataForType
+
+namespace browsing_data_test_util {
+
+void SetDataForType(const std::string& type,
+                    content::WebContents* web_contents) {
+  SetDataForType_ChromiumImpl(type, web_contents);
+  // Round-trip through RestrictedCookieManager to synchronise the CookieStore.
+  EXPECT_TRUE(HasDataForType(type, web_contents));
+}
+
+void SetDataForType(const std::string& type,
+                    content::RenderFrameHost* render_frame_host) {
+  SetDataForType_ChromiumImpl(type, render_frame_host);
+}
+
+}  // namespace browsing_data_test_util
diff --git a/test/filters/browser_tests.filter b/test/filters/browser_tests.filter
index eb676dfe5e9..2472c78a603 100644
--- a/test/filters/browser_tests.filter
+++ b/test/filters/browser_tests.filter
@@ -1944,10 +1944,12 @@
 # `kActionSendSharedTabGroupFeedback` ephemerally and triggers a CHECK.
 -VerticalTabViewDataSharingEnabledTest.LogsTabSwitchMetrics
 
-# Chromium test flaky - disabled on Windows upstream (crbug.com/413259587). Race
-# condition between JavaScript cookie set via document.cookie and backend
-# SiteDataCountingHelper query. Brave has no modifications to this code.
-# Upstream flake rate 0.5% (LUCI Analysis, 30-day lookback); amplified by MSAN.
+# Chromium test: BrowsingDataRemover BrowserContext destruction hangs in Brave.
+# BlockUntilCompletion() never returns after incognito profile is destroyed
+# during data removal (crbug.com/413259587).  Also hits a cookie IPC race
+# (GetSiteDataCount returns 0) that is fixed by the chromium_src override of
+# browsing_data_test_util.cc, but the hang remains.  Upstream flake rate 0.6%
+# (LUCI Analysis, 30-day lookback).
 -All/BrowsingDataRemoverBrowserTestP.BrowserContextDestructionVsCookieRemoval/*
 -All/BrowsingDataRemoverBrowserTestP.CookieIncognitoDeletion/*