Getting a table out of a PDF and into Excel without retyping it

Retyping a forty-row table from a PDF statement is nobody's idea of an afternoon. Conversion can usually spare you — if you know which tables convert cleanly and how to fix the ones that arrive messy.

Last updated October 2026

Conversion · 6 min read

Why tables are the hardest thing to convert

A PDF stores no table structure at all — no columns, no rows, no cells. It stores characters at coordinates. Conversion software looks at the geometry of those coordinates and infers the grid: characters that share a horizontal band form a row, and clusters with similar x-positions form columns.

That inference works well when the table was drawn with real rules or aligned precisely, and poorly when it was not. This is why one PDF's table converts perfectly and the next one's collapses into soup.

Tables that convert cleanly

Tables with ruled lines or borders — the strongest signal of all.
Financial statements and invoices from real systems, which align columns to a grid.
One table per page, or one dominant table with a clear title above it.
Digital PDFs — a text layer is mandatory. A scanned table has no characters to place at all.

What you get, realistically

Expect the numbers, dates and labels to arrive correctly positioned. Expect some cosmetic mess: merged cells that split into repeated values, currency symbols in their own column, totals rows sitting in odd places, and multi-line cell text arriving as two rows.

Treat conversion as producing a well-structured draft, not a finished workbook. The cleanup below is faster than retyping by an order of magnitude, but it is not zero work.

Cleaning up in Excel, fast

Select everything, set vertical alignment to top and disable wrap text — most 'broken rows' reveal themselves immediately.
Use Text to Columns on any column where two fields landed together (date plus description, code plus name).
TRIM the results to remove stray spaces that break lookups and comparisons.
Convert text-formatted numbers with the little green warning triangles' bulk-convert option before doing any arithmetic.
Sort a sample column ascending and descending to spot rows that belong together but arrived apart.

When retyping is actually faster

Small tables — a handful of rows — are often quicker to type than to convert and clean. The same is true for tables spanning awkward page breaks with repeated headers, and for anything where you need to restructure the data anyway rather than preserve it.

For big, clean tables the economics always favour conversion. For a 3×4 grid, just type it.

Tools mentioned in this guide

Keep reading

All guides