Translating a PDF seems straightforward: extract the text, run it through a translation engine, put it back. But PDFs store text in positioned blocks, not flowing paragraphs. Translated text is almost always longer or shorter than the original. A German translation of an English document expands by 20 to 30 percent. A Chinese translation contracts. The translated text no longer fits the spaces designed for the original language. The result is overflowing text boxes, broken layouts, and a document that looks like it was assembled by an algorithm that has never seen a page.
Preserving formatting during PDF translation requires understanding which elements of a PDF will survive the translation process and which will need manual adjustment afterward. This guide covers the translation workflow that maximizes what is preserved and clearly identifies what will need your attention.
According to a 2025 report by the language services industry group Nimdzi, demand for document translation grew 34 percent between 2022 and 2025, with PDFs accounting for over 60 percent of translated business documents (Nimdzi, "Global Language Services Market Report," 2025). The volume is growing. The formatting challenge is not.
The Translation Workflow That Preserves the Most Formatting
The highest-fidelity translation workflow does not translate the PDF directly. It extracts the text, translates it, and then rebuilds the document. Start by converting the PDF to an editable format like Word. The conversion preserves the document structure: headings, paragraphs, tables, and images all transfer to the editable file. Translate the Word document using a translation tool or service. Then review and adjust the formatting in Word before converting back to PDF.
This workflow preserves more formatting than direct PDF translation because Word handles text reflow natively. When the translated text expands, Word pushes content to the next line or page automatically. PDF editors do not reflow text. They place characters at fixed positions. The Translate PDF workflow through an editable format avoids the fixed-position constraint entirely.
When Direct PDF Translation Is the Only Option
If the PDF cannot be converted, because it contains scanned images rather than text or because the layout is too complex for conversion to handle, direct translation is the fallback. Browser-based PDF translation tools extract the text layer, translate it, and overlay the translation onto the original document. The layout is preserved because the underlying page structure does not change. But the translated text will almost certainly overflow or underflow its original space.
After direct translation, review every page at 100 percent zoom. Look for text that extends beyond its container, overlaps with adjacent elements, or is cut off mid-sentence. These issues are unavoidable in direct PDF translation and require manual adjustment. The PDF Format challenge is that translated text and original layout are in fundamental tension. Accept that some manual cleanup will be needed.
Handling Text Expansion and Contraction
Text expansion varies by language pair. English to German typically expands 20 to 30 percent. English to Spanish expands 15 to 25 percent. English to French expands 15 to 20 percent. English to Chinese contracts 20 to 40 percent. Knowing the expected expansion or contraction for your language pair helps you anticipate where formatting problems will occur. Dense paragraphs are where expansion causes the most damage. Wide-open layouts with generous whitespace are more forgiving.
WukongPDF's PDF Tools platform supports the extract-translate-rebuild workflow. Convert the PDF to Word, translate the content, and convert back. The formatting survives because you control it at every step rather than trusting an automated process to get it right.
Reviewing Translated PDFs Before Delivery
A translated PDF needs three reviews. The language review checks that the translation is accurate and natural. The layout review checks that no text is missing, overlapping, or cut off. The functional review checks that links still work, form fields are still fillable, and bookmarks still point to the right pages. Each review catches errors the others miss. Skipping any of them risks delivering a document that is linguistically correct but practically unusable.
If you do not speak the target language, have a native speaker review the translation. Machine translation quality has improved dramatically but still produces errors that a non-speaker cannot catch. A five-minute review by a native speaker catches the errors that would embarrass you in front of a client or partner.