diff --git a/postgres/cds-plugin.js b/postgres/cds-plugin.js
index b91593583..7213c86db 100644
--- a/postgres/cds-plugin.js
+++ b/postgres/cds-plugin.js
@@ -65,15 +65,80 @@ cds.build?.register?.('postgres', class PostgresBuildPlugin extends cds.build.Pl
 
     promises.push(this.write(cds.compile.to.json(model)).to(path.join('db', 'csn.json')))
 
-    let data
-    if (fs.existsSync(path.join(this.task.src, 'data'))) {
-      data = 'data'
-    } else if (fs.existsSync(path.join(this.task.src, 'csv'))) {
-      data = 'csv'
-    }
-    if (data) {
-      promises.push(this.copy(data).to(path.join('db', 'data')))
+    // csvFileDetection (default on, for parity with the HANA build task) collects initial
+    // data from all model sources - app-local db/data AND every reuse module's db/data - so
+    // the deployer artifact matches `cds deploy` from source. When disabled, only the
+    // app-local data/csv folder is copied (legacy behaviour).
+    const csvFileDetection = this.task.options?.csvFileDetection ?? true
+    if (csvFileDetection && typeof cds.deploy?.resources === 'function') {
+      for (const [dest, sources] of await collectInitialData(model)) {
+        const to = path.join('db', 'data', dest)
+        if (sources.length === 1) {
+          promises.push(this.copy(sources[0]).to(to))
+        } else {
+          promises.push(this.write(mergeCsvFiles(sources)).to(to))
+        }
+      }
+    } else {
+      let data
+      if (fs.existsSync(path.join(this.task.src, 'data'))) {
+        data = 'data'
+      } else if (fs.existsSync(path.join(this.task.src, 'csv'))) {
+        data = 'csv'
+      }
+      if (data) {
+        promises.push(this.copy(data).to(path.join('db', 'data')))
+      }
     }
     return Promise.all(promises)
   }
 })
+
+// Discover initial-data resources across app + reuse modules, grouped by their
+// destination file name. Files that resolve to the same destination (e.g. a code list
+// extended by a reuse module and the consumer) are merged; init.js/ts is skipped as it
+// cannot be reproduced from the artifact's db/data folder.
+async function collectInitialData (model) {
+  const resources = await cds.deploy.resources(model)
+  const byDest = new Map()
+  for (const [file, entity] of Object.entries(resources)) {
+    if (entity === '*') continue // init.js/ts
+    const dest = path.basename(file)
+    const group = byDest.get(dest)
+    if (group) group.push(file)
+    else byDest.set(dest, [file])
+  }
+  return byDest
+}
+
+// Merge several CSV files for the same entity into the union of their rows under a unified
+// header. Files are discovered reuse-module-first, so base rows precede consumer rows.
+function mergeCsvFiles (files) {
+  const columns = []
+  const seen = new Set()
+  const records = []
+  for (const file of files) {
+    const [header, ...rows] = cds.parse.csv(fs.readFileSync(file, 'utf8'))
+    if (!header) continue
+    for (const column of header) if (!seen.has(column)) { seen.add(column); columns.push(column) }
+    for (const row of rows) {
+      const record = {}
+      header.forEach((column, i) => { record[column] = row[i] })
+      records.push(record)
+    }
+  }
+  const lines = [columns.map(csvEscape).join(',')]
+  const emitted = new Set()
+  for (const record of records) {
+    const line = columns.map(column => csvEscape(record[column] ?? '')).join(',')
+    if (emitted.has(line)) continue
+    emitted.add(line)
+    lines.push(line)
+  }
+  return lines.join('\n') + '\n'
+}
+
+function csvEscape (value) {
+  value = String(value)
+  return /[",\n\r]/.test(value) ? '"' + value.replace(/"/g, '""') + '"' : value
+}
diff --git a/postgres/test/cds-build.test.js b/postgres/test/cds-build.test.js
index 8d897581d..af009e2b1 100644
--- a/postgres/test/cds-build.test.js
+++ b/postgres/test/cds-build.test.js
@@ -12,9 +12,13 @@ const genDir = path.join(workDir, 'gen')
 const pgDest = path.join(genDir, 'pg')
 const dbDest = path.join(pgDest, 'db')
 
+const reuseWorkDir = path.join(__dirname, 'reuse-sample')
+const reuseGenDir = path.join(reuseWorkDir, 'gen')
+
 // delete the generated folder after each test
 afterEach(() => {
   if (fs.existsSync(genDir)) fs.rmSync(genDir, { recursive: true })
+  if (fs.existsSync(reuseGenDir)) fs.rmSync(reuseGenDir, { recursive: true })
 })
 
 describe('cds build plugin', () => {
@@ -61,4 +65,12 @@ describe('cds build plugin', () => {
     const pgAdapterVersion = require(path.join(__dirname,'..','package.json')).version;
     expect(packageJson.dependencies?.['@cap-js/postgres']).to.equal(pgAdapterVersion)
   })
+
+  test('should merge initial data of reuse modules into gen/pg/db/data', () => {
+    execSync('npx cds build --for postgres', { cwd: reuseWorkDir })
+    const csv = fs.readFileSync(path.join(reuseGenDir, 'pg', 'db', 'data', 'my.reuse-CodeList.csv'), 'utf8')
+    const codes = csv.trim().split(/\r?\n/).slice(1).map(line => line.split(',')[0])
+    // union of the reuse module's rows (A, B, C) and the consumer's own row (D)
+    expect(codes.sort()).to.eql(['A', 'B', 'C', 'D'])
+  })
 })
diff --git a/postgres/test/reuse-sample/db/data/my.reuse-CodeList.csv b/postgres/test/reuse-sample/db/data/my.reuse-CodeList.csv
new file mode 100644
index 000000000..e518296d6
--- /dev/null
+++ b/postgres/test/reuse-sample/db/data/my.reuse-CodeList.csv
@@ -0,0 +1,2 @@
+code,name
+D,Delta
diff --git a/postgres/test/reuse-sample/db/index.cds b/postgres/test/reuse-sample/db/index.cds
new file mode 100644
index 000000000..0198e96c6
--- /dev/null
+++ b/postgres/test/reuse-sample/db/index.cds
@@ -0,0 +1,3 @@
+// The reuse module is referenced via a relative path so the fixture needs no
+// node_modules; resource discovery treats its db/data folder like any dependency's.
+using from '../reuse/db/schema';
diff --git a/postgres/test/reuse-sample/package.json b/postgres/test/reuse-sample/package.json
new file mode 100644
index 000000000..52d3a92ba
--- /dev/null
+++ b/postgres/test/reuse-sample/package.json
@@ -0,0 +1,15 @@
+{
+  "name": "reuse-sample",
+  "version": "1.0.0",
+  "description": "A CAP project consuming a reuse module, to test the build plugin",
+  "dependencies": {
+    "@cap-js/postgres": "../../."
+  },
+  "cds": {
+    "requires": {
+      "db": {
+        "kind": "postgres"
+      }
+    }
+  }
+}
\ No newline at end of file
diff --git a/postgres/test/reuse-sample/reuse/db/data/my.reuse-CodeList.csv b/postgres/test/reuse-sample/reuse/db/data/my.reuse-CodeList.csv
new file mode 100644
index 000000000..2723a2108
--- /dev/null
+++ b/postgres/test/reuse-sample/reuse/db/data/my.reuse-CodeList.csv
@@ -0,0 +1,4 @@
+code,name
+A,Alpha
+B,Beta
+C,Gamma
diff --git a/postgres/test/reuse-sample/reuse/db/schema.cds b/postgres/test/reuse-sample/reuse/db/schema.cds
new file mode 100644
index 000000000..db7620086
--- /dev/null
+++ b/postgres/test/reuse-sample/reuse/db/schema.cds
@@ -0,0 +1,6 @@
+namespace my.reuse;
+
+entity CodeList {
+  key code : String;
+  name     : String;
+}
diff --git a/postgres/test/reuse-sample/reuse/package.json b/postgres/test/reuse-sample/reuse/package.json
new file mode 100644
index 000000000..6260a25f5
--- /dev/null
+++ b/postgres/test/reuse-sample/reuse/package.json
@@ -0,0 +1,5 @@
+{
+  "name": "@cap-js-test/reuse-code-list",
+  "version": "1.0.0",
+  "description": "A reuse module shipping initial data, to test the build plugin"
+}
\ No newline at end of file