How to fill in a PDF form that will not let you type

Work out which of the three you have

Open the file and try two things.

First, try to select the text. Drag across a printed label such as "Full name". If the label highlights, the document has a text layer and was produced digitally. If nothing highlights and you get a selection rectangle over the whole page, the page is an image: a scan or a photograph. Is my PDF scanned or text? goes through this in more detail.

Then click where an answer should go. A real interactive form puts a cursor in a box, and most readers tint the fillable boxes pale blue when the file opens. No cursor and no tint means there are no fields, whatever the page looks like.

That gives you the answer:

• No text selects, no fields — a scan. The next section.
• Text selects, no fields — a flattened or never-interactive form. Also the next section; the handling is the same.
• Fields are visible but refuse input — protected. Two sections down.
• A message about needing a different reader — an XFA form. The section after that.

A picture of a form: put your answers on top

This covers the great majority of cases, and the thing to accept early is that there is nothing to repair. The document has no fields to recover, because it never had any. What you do instead is add your own text at the right positions on the page.

Edit PDF places text boxes anywhere on a page. For a form, the work is almost entirely about getting the alignment right, so a few practical notes:

• Match the line, not the label. Put the baseline of your text on the printed rule, not level with the question. Text floating a few millimetres high is the single thing that makes a filled form look wrong.
• Pick one size and keep it. Answers set at four different sizes read as though four people filled it in.
• Checkboxes. A letter x in the box is accepted everywhere and survives printing, faxing and re-scanning. A drawn tick often does not.
• Do the signature last. Sign PDF places a signature image or a drawn signature; how to sign a PDF covers what a signature on a PDF does and does not prove.

One thing worth not doing: running the scan through OCR to "make it fillable". Recognising text turns the picture of a word into a word, which is useful for searching and copying — but it does not create form fields, because there were none to find. OCR and PDF to Word are different jobs explains the distinction.

A real form that refuses input

If the fields are clearly there — they tint, they have borders, perhaps some already contain text — and typing does nothing, one of three things is true.

It was flattened. Flattening paints what the fields were showing permanently into the page and then deletes the fields, which is exactly what you do to a form once it is finished. A flattened form looks complete because it is complete; somebody sent you the finished version rather than the blank one. There is no undo, and the honest answer is to ask for the original. What flattening a PDF means covers what it changes.

It is protected. A PDF can carry a permissions setting that allows reading but forbids filling in fields. Your reader is obeying it. This is a restriction the author set, so the right first step is to ask them for a version without it rather than to work around it.

Your reader cannot do forms. Browsers' built-in PDF viewers have become good at this and some phone readers still have not. Before concluding the file is at fault, open it somewhere else. Fill in a PDF form lists the fields a document actually declares, which is a quick way to settle the question: if it finds fields, they exist and the problem is the reader.

The forms that ask for a different reader

A small number of forms — government and banking applications, mostly, produced in the 2000s by Adobe LiveCycle — are not really PDFs. They are XFA forms wearing a PDF wrapper, and the wrapper usually contains a single page saying "please open this with Adobe Reader".

There is no trick here. The form's structure is in an XML layer that almost nothing outside Adobe's own desktop reader implements; browsers, phone readers and third-party tools show the warning page because it is the only part of the file they can render. Adobe itself ended support for the technology, so it is a shrinking problem rather than a growing one.

Three things that do work, roughly in order of how little time they cost: open it in Adobe Acrobat Reader on a desktop, which is free; ask the issuer whether they have an ordinary PDF or a web form, which many now do; or print it, complete it by hand, and scan it back. What does not work is converting it — the conversion produces the warning page, because that is the page the file shows to anything that is not Adobe Reader.

Before you send it back

Three checks, each of which catches a different way a completed form arrives wrong.

Flatten it, if it had real fields. Flatten PDF fixes your answers into the page. Until you do, the values live in field objects, and a reader that renders fields differently — or a recipient who clicks into one — can change or lose them. Flattening also removes any possibility of the next person editing what you wrote.

Open it somewhere you did not fill it in. A browser if you used a desktop reader, a phone if you used a browser. A form that looks right only in the application that produced it is the commonest avoidable failure, and it takes ten seconds to rule out.

Look at the document properties. A form you received carries the author, the software and sometimes a name from the organisation that produced it, and your answers do not replace any of that. If the completed form is going somewhere sensitive, remove the metadata first; what PDF metadata contains lists what is in there.