Convert DOCX to CSV

A database, spreadsheet, contact manager, or data-processing script may accept tabular data only as CSV and reject a Word document. Converting the table to CSV makes its rows and fields importable, but it does not preserve the document's visual layout.

What the DOCX format is

DOCX is the XML-based document format used by Microsoft Word and supported by LibreOffice Writer, Apple Pages, Google Docs, and document-management systems. It can contain formatted paragraphs, tables, images, headers, footers, comments, and other layout information. DOCX files commonly contain reports, forms, invoices, and lists.

A DOCX file can contain structured tables as well as ordinary paragraphs. A table can usually be exported as rows and columns; text that is merely aligned with spaces or tabs needs to be structured first. The older .doc format is different internally, but its tables can also be copied into a spreadsheet and saved as CSV.

What the CSV format is

CSV, or comma-separated values, is plain text in which records normally occupy lines and fields are separated by a delimiter, usually a comma. Spreadsheet applications, databases, command-line tools, statistical software, and programming languages can read it without requiring page-layout information.

CSV is suitable for filtering, sorting, importing, archiving, and transferring structured records. It cannot preserve fonts, images, page breaks, comments, merged-cell appearance, or most other document formatting. A CSV file is normally one rectangular table; unrelated tables from the same DOCX should usually be exported separately.

How to convert one table with desktop software

1. Open the DOCX in Microsoft Word or LibreOffice Writer and confirm that the source is an actual table. If it is plain text, convert it to a table first using Table → Convert → Text to Table in Word, or the equivalent Writer command.
2. Select the table and copy it. In Writer, use Table → Select → Table before copying.
3. Paste it into cell A1 in Microsoft Excel or LibreOffice Calc.
4. Remove title rows, notes, blank lines, and decorative columns that should not become imported fields. Check that every record occupies one row and every field one column.
5. In Excel, choose File → Save As and select CSV UTF-8 (Comma delimited) (*.csv). In Calc, choose File → Save As, select Text CSV (.csv), and set the character set to UTF-8.
6. When Calc displays the export options, choose comma as the field delimiter unless the destination specifies another delimiter, and keep text quoting enabled.

Excel and Calc may warn that CSV cannot preserve workbook features; confirm the CSV format. Word does not generally provide a direct DOCX-to-CSV export command, so copying the table through a spreadsheet is the practical manual method.

How to automate table extraction

For repeated conversions, install python-docx with python -m pip install python-docx. Save this as docx_table_to_csv.py to export the first table:

from csv import writer
from docx import Document

source = Document("input.docx")
if not source.tables:
    raise ValueError("No table found")

table = source.tables[0]
with open("output.csv", "w", newline="", encoding="utf-8-sig") as file:
    csv_writer = writer(file)
    for row in table.rows:
        csv_writer.writerow([cell.text.replace("\n", " ").strip() for cell in row.cells])

Place input.docx beside the script and run python docx_table_to_csv.py. Change source.tables[0] to another index for a different table. For multiple tables, write each table to a separate CSV unless they have identical headers and columns; appending unrelated tables to one CSV produces an ambiguous dataset.

Online conversion

CloudConvert, Convertio, and Zamzar may offer DOCX-to-CSV conversion, although available formats and extraction behavior can change. Upload only non-confidential files, verify that the service explicitly supports this conversion, inspect the downloaded CSV, and delete the upload when the service provides that option.

Online services may extract visible text instead of table cells, mishandle multiple tables, or fail on nested tables and complex layouts. Desktop extraction is preferable for confidential documents, large batches, or repeatable results.

Quality and compatibility checks

• There is no universal DOCX-to-CSV layout mapping. A Word page is document-oriented, while CSV represents fields and records in a rectangular structure.
• Merged cells, nested tables, repeated header rows, footnotes, and line breaks inside cells require manual review. Extractors may duplicate merged values, create empty fields, or flatten line breaks.
• Fields containing commas, quotation marks, or line breaks must be enclosed in double quotes according to CSV escaping rules. Do not remove those quotes during cleanup.
• Use UTF-8 for accented characters and non-Latin scripts. A UTF-8 BOM, produced by utf-8-sig, can improve recognition when the file opens directly in some older Excel versions, but it is unnecessary for systems that expect plain UTF-8.
• Delimiter conventions vary by application and locale. A receiving system may require semicolons or tabs instead of commas; use its specified delimiter and confirm whether it expects a header row.
• Excel can reinterpret identifiers such as 00123, dates, long numbers, and account codes. Import the CSV through Excel's data-import dialog and set identifier columns to Text rather than opening the file by double-clicking.
• Compare the CSV's headers, row count, column count, and representative records with the DOCX before importing it into another system. Check the output in both a plain-text editor and a spreadsheet when quoting or encoding is in doubt.