How to turn a PDF table into an Excel spreadsheet

From a page to usable rows

1. Open the PDF table extractor and choose a PDF, or try its built-in example. Keep an unchanged copy of your original.
2. In the original-page preview, choose the page containing your table. Drag a rectangle around the complete table, including its column headings. Leave a little space around the text: do not cut through a word or value.
3. Select Extract this selection. Compare the rows with the page beside them. If the area is wrong, draw a new rectangle and extract again. Each region applies to one page.
4. Choose Download Excel (.xlsx) for a workbook, or Download CSV files (.zip) for separate CSV files. Only supported, selected tables enter these spreadsheet downloads.
5. To paste into an existing spreadsheet, choose a table from the dropdown and select Copy table. Check the pasted columns, leading zeros and row count before continuing.

Want several complete pages instead? Select Clear region, leave Pages blank for all pages or enter a range such as 2-4, then select Extract tables. Uncheck unwanted tables and use Update downloads with selected tables.

Put a table spanning multiple PDF pages into one sheet

1. Select Clear region if an area is drawn. Enter the relevant pages, or leave Pages blank, and select Extract tables. The built-in Matching multi-page table is a safe way to try this without a private document.
2. Compare both pages with the extracted rows. Check that the tables belong to the same dataset and their column headings and order match. Uncheck unwanted tables and select Update downloads with selected tables.
3. Expand Combine matching tables into one sheet and choose Download combined Excel or Download combined CSV. At least two supported tables are required. Mismatched headings, reordered columns or inconsistent row widths refuse the combined export; use separate sheets instead.
4. Check that the combined data-row count equals the sum of the selected tables. The header is written once. Every data row, including duplicates, is retained in extraction order. Rows that exactly equal the headings are kept and reported for review; they are not guessed away.
5. Use the extra source-page and source-table columns to locate each value in the PDF. Their names receive a numbered suffix if an original column already uses that name. Excel keeps cells as text; import CSV IDs and date-like values as Text to prevent spreadsheet conversion.

The original Excel download remains one sheet per table. Combining does not alter those files or the extraction receipt. Combined export details (.json) records the chosen source tables, row count and hashes of the two derived files, linked to the original receipt. It does not certify that values are correct.

Appending tables is a standard multi-page extraction task: Microsoft documents its MultiPageTables option. FastTool uses an explicit, narrower rule: matching headers and widths, with no fuzzy mapping, sorting, type inference or deduplication.

Decision 1: is this a supported digital-text PDF?

Open the source in an independent viewer and try selecting a short cell value. Selection alone is not proof, but it helps distinguish digital text from an image-only scan. Use the extractor only when the file is authorized, begins with a normal PDF signature, is no larger than 8 MiB and exposes no more than 50 page objects.

• Encrypted: stop. The tool returns BLOCK and does not attempt decryption.
• Scanned or image-only: stop the extraction claim. The tool returns INDETERMINATE because it performs no OCR.
• Rotated, skewed, unresolved-font or unsupported stream: retain the coverage warning and use manual review or another bounded workflow.
• Digital text with stable coordinate rows: continue, but still compare the result with the page.

Make the selection precise without a mouse

Select Select whole visible page to start from the page boundary. Focus the page area, then use arrow keys to move the selection. Hold Shift for larger steps, or Alt with an arrow to resize. Escape clears it. For exact values, expand Exact region coordinates, enter the left, bottom, width and height in PDF points, then select Apply region coordinates.

Zoom and Rotate view 90° change how you inspect the page, not the document or its extraction support. A PDF containing rotated or unsupported text can still be refused. Changing the page clears its previous region. Every selection or extraction-setting change locks older downloads until you run the extraction again.

If two columns become one

First select a smaller area containing only one table. Include the headings and at least two data rows; exclude nearby footnotes, page numbers and other tables. Run the extraction again and compare the result.

If the columns still need help, set vertical separators in the tool’s extraction settings. Values such as 200,400 are positions in the original PDF coordinate system, measured in PDF points; they are not browser pixels or percentages. A separator divides the page into column bands. If it crosses text, or the layout remains ambiguous, do not force the output into a spreadsheet that merely looks plausible. Correct the boundary or use another extraction method.

Decision 3: read PASS, BLOCK and INDETERMINATE literally

Excel, CSV or copy: which should you choose?

• Excel (.xlsx): the easiest option for several tables. Each table gets its own sheet named with its source page, original columns, a frozen header and column filters.
• CSV files (.zip): one rectangular CSV per selected table. Unzip the download and import the file you need. Different table layouts are never silently combined.
• Copy table: tab-separated rows for pasting into Excel or Google Sheets. Choose the table you want before copying.
• Cell log (.csv), under More formats: an optional audit list with source page, row and column for each cell. This is not the normal spreadsheet export.
• JSON, map, notes and receipt: optional review files containing raw values, coordinates, extraction limits and exact file hashes. The map is not a rendered PDF page; matching hashes establish file identity, not correct interpretation.

The Excel (.xlsx) export explicitly stores cells as text, preserving leading zeros without guessing whether 03/04 is a date. CSV files and copied rows contain plain text but cannot enforce a spreadsheet cell type: Excel or Sheets may automatically change IDs and dates. Choose Text for those columns in the import dialog, or format the destination columns as text before pasting. After checking the original, convert only the intended columns to numbers or dates. Formula-leading values are protected with an apostrophe.

Check these before using the spreadsheet

1. Compare the first and last row, then every value you intend to use, against the original page. Make sure a title or footnote has not become a data row.
2. Check minus signs, decimal separators, dates, currency labels, blank cells and totals separately. A correctly shaped table can still contain an incorrect interpretation.
3. Count the rows and confirm the selected page and table scope. Unrelated tables should stay separate, even when their column headings match.
4. Open the downloaded file in your spreadsheet application. Confirm IDs such as 00123 stay unchanged before converting columns to numbers.

For a review trail, the optional receipt and cell log record the input and extraction settings. They help identify files; they do not prove the values mean what you think they mean. For consequential financial, legal or medical use, obtain appropriate human review.

Why are the columns wrong, or the download unavailable?

• Everything appears in one column: when opening CSV, use your spreadsheet's import dialog and select comma as the separator. Or choose XLSX to avoid delimiter settings.
• Rows do not line up: compare the preview with the PDF. Try a smaller page selection or manual column boundaries. The tool cannot reliably fix merged cells or complex page layouts.
• The PDF is a scan: this tool does not perform OCR. Use a separate OCR workflow and check its recognition errors before relying on the result.
• Numbers will not sum: values intentionally remain text. Confirm decimal separators, units and leading zeros, then convert the appropriate column in your spreadsheet.
• A setting changed: downloads are locked until you extract again. This prevents saving a result from the previous file or selection.

Sources, correction and review date

Multi-page combining workflow updated 17 October 2026; the existing PDF extraction boundaries are unchanged. Primary references: PDF / ISO 32000 resources, RFC 4180, OWASP CSV Injection, and ECMA-376.

Report an incorrect result or unsupported claim to contact@fasttool.app. Include the behavior and steps to reproduce it; never email confidential source data.