diff --git a/account_statement_import_sheet_file_bg/models/__init__.py b/account_statement_import_sheet_file_bg/models/__init__.py
index ae69bca2..fb0170f6 100644
--- a/account_statement_import_sheet_file_bg/models/__init__.py
+++ b/account_statement_import_sheet_file_bg/models/__init__.py
@@ -1 +1,2 @@
 from . import account_statement_import
+from . import account_statement_import_sheet_parser
diff --git a/account_statement_import_sheet_file_bg/models/account_statement_import.py b/account_statement_import_sheet_file_bg/models/account_statement_import.py
index 8c869b3d..8fb99cde 100644
--- a/account_statement_import_sheet_file_bg/models/account_statement_import.py
+++ b/account_statement_import_sheet_file_bg/models/account_statement_import.py
@@ -32,24 +32,22 @@ class AccountStatementImport(models.TransientModel):
     def import_file_button(self, wizard_data=None):
         """Process the file chosen in the wizard, create a bank statement
         and return a link to its reconciliation page."""
-        if not self._context.get("bg_job"):
-            if self.sheet_mapping_id:
-                header_column = self.sheet_mapping_id.header_lines_skip_count
-                # Get row limit from system parameter
-                rows_limit = (
-                    self.env["ir.config_parameter"]
-                    .sudo()
-                    .get_param("account_statement_import_sheet_file_bg.rows_per_file_limit")
-                )
-                # Only split if parameter exists and has a valid value
-                files = []
-                if rows_limit:
-                    try:
-                        rows_limit = int(rows_limit)
-                        files = self.split_base64_excel(header_column, rows_limit)
-                    except (ValueError, TypeError):
-                        files = []
+        rows_limit = (
+            self.env["ir.config_parameter"]
+            .sudo()
+            .get_param("account_statement_import_sheet_file_bg.rows_per_file_limit")
+        )
+        if not self._context.get("bg_job") and rows_limit:
+            # Validate parameter is a valid integer
+            try:
+                rows_limit = int(rows_limit)
+            except (ValueError, TypeError):
+                # Parameter not valid, skip bg processing
+                rows_limit = None
 
+            if rows_limit and self.sheet_mapping_id:
+                header_column = self.sheet_mapping_id.header_lines_skip_count
+                files = self.split_base64_excel(header_column, rows_limit)
                 if files:
                     for idx, file in enumerate(files):
                         # Encode the file to string format, because background jobs cannot
@@ -171,7 +169,6 @@ def split_base64_excel(self, header_rows_count, rows_per_file_limit):
         mapping = self.sheet_mapping_id
         journal = self.env["account.journal"].browse(self.env.context.get("journal_id"))
         currency_code = (journal.currency_id or journal.company_id.currency_id).name
-
         try:
             file_bytes = base64.b64decode(self.statement_file)
             read_buffer = BytesIO(file_bytes)
@@ -179,7 +176,9 @@ def split_base64_excel(self, header_rows_count, rows_per_file_limit):
             # Try openpyxl (xlsx)
             input_workbook = load_workbook(read_buffer)
             input_worksheet = input_workbook.active
-            all_rows = list(input_worksheet.rows)
+            # Normalize rows to plain values to keep parser/output logic
+            # consistent with CSV/xls flows.
+            all_rows = [[cell.value for cell in row] for row in input_worksheet.rows]
             csv_or_xlsx = (input_workbook, input_worksheet)
 
         except Exception:
@@ -277,10 +276,14 @@ def _filter_rows_with_date(self, data_rows, date_column_index):
 
         filtered_rows = []
         for row in data_rows:
+            date_value = None
+            if len(row) > date_column_index:
+                date_cell_or_value = row[date_column_index]
+                date_value = date_cell_or_value.value if hasattr(date_cell_or_value, "value") else date_cell_or_value
             # Check if the row has enough columns and the date column is not empty
-            if len(row) > date_column_index and row[date_column_index].value:
+            if len(row) > date_column_index and date_value:
                 filtered_rows.append(row)
-            elif len(row) > date_column_index and not row[date_column_index].value:
+            elif len(row) > date_column_index and not date_value:
                 # Stop processing when we find the first empty date
                 break
 
diff --git a/account_statement_import_sheet_file_bg/models/account_statement_import_sheet_parser.py b/account_statement_import_sheet_file_bg/models/account_statement_import_sheet_parser.py
new file mode 100644
index 00000000..cfb631a6
--- /dev/null
+++ b/account_statement_import_sheet_file_bg/models/account_statement_import_sheet_parser.py
@@ -0,0 +1,31 @@
+# Copyright 2026 ADHOC SA
+# License AGPL-3.0 or later (https://www.gnu.org/licenses/agpl).
+
+from odoo import api, models
+
+
+class AccountStatementImportSheetParser(models.TransientModel):
+    _inherit = "account.statement.import.sheet.parser"
+
+    @api.model
+    def parse_header(self, csv_or_xlsx, mapping):
+        if mapping.no_header:
+            return []
+
+        header_line = mapping.header_lines_skip_count
+        # Prevent negative indexes.
+        if header_line > 0:
+            header_line -= 1
+
+        if isinstance(csv_or_xlsx, tuple):
+            return super().parse_header(csv_or_xlsx, mapping)
+
+        [next(csv_or_xlsx) for _i in range(header_line)]
+        header = []
+        for value in next(csv_or_xlsx):
+            raw_value = value.value if hasattr(value, "value") else value
+            header.append(str(raw_value).strip() if raw_value is not None else "")
+
+        if mapping.offset_column:
+            header = header[mapping.offset_column :]
+        return header