How to Print Markdown to PDF With Tables That Stay Readable

Short answer: GFM tables + a renderer that keeps rows whole

Markdown tables are a GitHub Flavored Markdown extension: one header row, a delimiter row, and body rows separated by pipes. Every serious exporter understands them. The differences show up when a table meets the edge of the page, either the right edge (too many columns) or the bottom edge (too many rows).

Three things decide whether a printed table stays readable: the width of the table compared with the page, what happens at a page break, and whether the result is real text. The rest of this guide takes them one at a time, with numbers from our own exporter measured on Oct 18, 2026.

Writing tables that print well

When you print Markdown to PDF with tables, most problems start in the source file. A few habits prevent them:

• Keep cells short. Put long explanations in a paragraph under the table and leave a short label in the cell.
• Use alignment for numbers. A ---: delimiter right-aligns a column, which makes figures easy to compare on paper.
• Always include a header row. GFM requires one, and it is the row readers look for when a table runs onto a second page.
• Avoid long unbroken strings. URLs and file paths cannot wrap at spaces, so they force a column wide. Shorten them or move them into a footnote-style list.
• Escape pipes in cells as \|, or the row gains an extra column.

| Task           | Owner | Status    |
| :------------- | :---- | --------: |
| Draft the brief | Dana  | Done      |
| Review tables  | Lee   | In review |
| Export the PDF | Sam   | Next      |

Before exporting, preview the table at roughly the printed width. In our exporter the A4 text column is about 680 CSS pixels wide. If it already scrolls sideways in a narrow preview, it will be cramped on paper.

Wide tables: options when columns don’t fit A4

Wide tables are where PDF exports most often fail quietly. When we started testing for this article, a 9-column table exported with our own tool had only 5 of 9 header cells fully on the page and none of the last column visible: the rest ran past the right edge. We changed the exporter the same day. Tables that would overflow now wrap their cells to fit, and in the repeat test all 9 columns were on the page with no text past the right margin.

Fitting is better than losing columns, but it is not pretty: very narrow cells break words. For wide tables PDF layout offers four better options, in this order:

1. Drop or merge columns that readers do not need on paper.
2. Split the table into two that share a key column, such as an ID or a name.
3. Transpose it so the many columns become rows.
4. Print in landscape. Our exporter writes A4 portrait only. Chromium-based exporters accept a CSS rule such as @page { size: A4 landscape; } (MDN: size).

Long tables across pages

A table split across pages has two failure modes: a row cut in half, with its top on one page and its bottom on the next, and a second page that starts with no header, so readers lose track of the columns. Both matter whenever you print Markdown to PDF with tables longer than a page. We tested them with a 60-row table whose rows each wrap onto two lines.

Our exporter keeps every row whole, but it does not repeat the header row on later pages. Chrome did both. The CSS 2.1 specification allows it without requiring it:

If you need repeated headers from our tool today, split a long table into page-sized parts and repeat the header row yourself. In HTML-based exporters, tr { break-inside: avoid; } (MDN) asks the browser to keep rows whole.

Print Markdown to PDF with tables: browser tool vs VS Code vs Pandoc

All three routes can print Markdown to PDF with tables. They differ in setup and in how much control you get.

For vscode markdown to pdf tables, the Markdown PDF extension documents a markdown-pdf.styles setting for custom CSS and a portrait or landscape option. It prints with a Chromium browser, the same engine as the Chrome row above, but we did not test the extension itself.

For pandoc markdown to pdf tables, the Pandoc manual says LaTeX output uses longtable by default, so tables “can span page boundaries”. It also says that when a pipe table has a line longer than the column width, relative column widths follow the number of dashes in the delimiter row. That gives you a way to make one column wider without HTML.

pandoc report.md -o report.pdf

Checking the PDF: selectable text, searchable cells

The last check when you print Markdown to PDF with tables is the text layer. A markdown table pdf export is only useful if the cells are real text. Some exporters, including the fallback mode of our own tool, save each page as an image inside the PDF, which looks fine but cannot be searched, copied or read by a screen reader. Check it in ten seconds: open the PDF, press Ctrl+F (or Cmd+F) and search for a word that appears only inside a table cell. Then select one row and paste it.

Our exporter writes real text with embedded font subsets. In our round-trip test, four tables printed to a PDF and converted back with our PDF to Markdown tool gave 17 of 17 rows identical to the source, including an escaped pipe and right-aligned numbers. For the other direction, see our guide on how to convert PDF tables to Markdown.