Guide

How to convert a scanned form into a fillable PDF

Updated 11 October 2026

Short answer: a scan is one image per page, so the fields have to be placed over the picture. Do it by hand for free in LibreOffice Draw, or use a tool that finds the lines and boxes in the image (Acrobat's Prepare a Form, Fillable) and then check every field it made. OCR alone won't do it: it makes the text searchable, not fillable.

Start with a clean scan

Flat and straight. A page fed through a scanner beats a phone photo. If you only have a photo, crop it to the page edges and keep it square on.
200 to 300 dpi, grayscale or black and white. Enough for thin lines to survive; colour at 600 dpi only makes the file huge.
One PDF, pages in order. Most tools work page by page, and a form split across files has to be done twice.

What a tool can find in a scan

Without text, software still sees shapes: long dark lines, small squares, big empty rectangles. Each one maps to a field kind.

Without text recognition a detector still sees shapes. What it can't read is the label, so fields from a scan need names.

Free: draw the fields over the scan in LibreOffice

Open the scanned PDF in LibreOffice Draw (File > Open). Each page comes in as an image, so the layout can't shift.
View > Toolbars > Form Controls, and switch on Design Mode.
Drag a Text Box over each blank and a Check Box over each square. Zoom to 200% so the fields line up with the printed lines.
Right-click each field, open Control Properties and give it a name.
File > Export as PDF, tick Create PDF form on the General tab, and export.

It's accurate because you place every field yourself, and it's slow for the same reason.

Do I need OCR first?

No. OCR (text recognition) turns the picture of words into selectable, searchable text. It doesn't create form fields. Adobe says Acrobat's Prepare a Form tool can work on a scan of a paper form and detect fields on it; independent guides add that the result needs checking, especially on complicated forms. Fillable doesn't do text recognition at all: it finds the lines and boxes in the image, so on a scanned page the fields start without names and you type them in.

With Fillable

Fillable's editor with a scanned intake form. A notice says 38 fields were found and that fields on a scanned page have no names yet. The field list shows Text 1, Text 2 and so on. — The same form as a 150 dpi scan, slightly tilted. The lines and boxes were found in the image; a few extra boxes were picked up around the labels in the middle, so check each field and delete those.

Open fillable.getnobody.org and choose the scanned PDF (up to 20 MB and 30 pages). Detection runs in your browser and the file isn't uploaded while you work.
Look over the outlines. Delete anything that isn't a blank, add what was missed from the toolbar (Text, Multiline, Checkbox, Date, Signature, Number).
Name the fields: select the first, type a name, press Enter to move to the next.
Download page 1 free and try it in your reader. It's the real fillable file with a small "Made with Fillable" line at the bottom. The whole form without the line is $4, once.

When it goes wrong

Problem	Likely cause	Fix
Fields sit slightly above or below the lines	The page was scanned at an angle	Rescan straight, or nudge the fields; they only need to cover the blank
Checkboxes missing	Faint or very small squares	Add them by hand; rescanning darker helps
Text box over a label	A boxed label looks like an empty box	Delete that field
File too large to open	Colour scan at high resolution	Scan again in grayscale at 200 to 300 dpi
"Password protected"	The PDF is encrypted	Open it, print it to a new PDF, and use that

Fillable finds the lines, boxes and checkboxes in your PDF in the browser, lets you fix them, and gives you back a real fillable PDF. Page 1 is free to download (with a small "Made with Fillable" line at the bottom), so you can try it in your own PDF reader. The whole form without that line is $4, once. No account, no subscription.

Make your form fillable

Sources

More: make any PDF fillable without Acrobat, PDF field types.