Crop Questions from a PDF: 4 Methods Compared (Honest Pros & Cons)

If you have searched for a way to pull individual questions out of a PDF, you have probably noticed the strange gap: the big PDF editors trim margins, the dedicated apps are phone-only, and everything else wants your file on a server.

This guide compares the four real methods for cropping questions out of a PDF — with the honest trade-offs of each — so you can pick one and stop researching.

The 4 methods at a glance

The rest of this guide explains why those numbers look the way they do.

Method 1 — Manual cropping in a PDF editor

Adobe Acrobat Pro, Preview on macOS, and most paid editors have a crop tool: open the page, drag the rectangle, apply, export.

Where it works: one or two questions, or when you are already living in Acrobat all day.

Where it breaks:

• It is one question per save. Cropping 30 questions means 30 crop operations and 30 save dialogs. The time cost is in the repetition, not the cropping.
• It does not understand page breaks. A question split across pages becomes two crops, and you have to stitch the halves into a new document yourself.
• It edits the page, not the question. Cropping pages in Acrobat removes content from the page geometry — easy to accidentally alter the source paper if you do not save a copy first.

Verdict: fine for a couple of questions. Wrong tool for a test book.

Method 2 — The iOS-only dedicated app

For years the only tool designed around “one question = one file” was an iPhone/iPad app. It proved the concept works: box each question, export each as its own PDF.

The limitations are structural:

• Platform. No Windows, no Android, no Chromebook — which describes most school computer labs.
• Screen size. Boxing a full-width question on a phone screen is fiddly, and a 40-question paper means a lot of scrolling.
• Bulk workflows. Getting files off the phone and into your question bank is an extra sync step.

Verdict: genuinely useful if you live entirely in the Apple ecosystem and work with small papers.

Method 3 — Generic “split PDF” web tools

Free web tools that split, merge, or extract pages are everywhere. But look at what they actually split on: pages. A test paper has one page per ~3 questions, so page-splitting gives you wrong-sized chunks that you then have to cut again.

And there is the upload problem. These tools take your file server-side — which is exactly what you do not want for an exam paper that has not been administered yet. See is it safe to upload a PDF to an online tool? for how to judge that risk.

Verdict: wrong unit of extraction, plus an upload you should not be making with confidential papers.

Method 4 — A browser-based question cropper

This is the category QuestionCropper was built for, and it is worth being specific about what makes it different rather than just faster:

• The unit is the question. You draw boxes; each box becomes one numbered PDF.
• Page breaks are handled. Drag one box from the bottom of page 1 into the top of page 2 and the two fragments are stitched into a single continuous question — the case every other method above fails.
• Nothing is uploaded. Rendering, cropping, and rebuilding all happen in your browser. Your Network tab shows zero file requests, which is the right property for a confidential exam. The full redaction-and-privacy argument is in how to redact a PDF permanently.
• One export, one ZIP. All 30 crops come down as Question-1.pdf … Question-30.pdf in a single download — no save dialogs.

The full walkthrough, including drawing tips, is in how to crop questions from a PDF test book.

Which should you use?

• 1–2 questions, already in Acrobat → Method 1. It is fine.
• Apple-only, small papers → the iOS app still works.
• Any real paper, any platform → the browser cropper. It is the only method where a 30-question paper takes minutes instead of an hour, and the only one that gets page-break questions right without manual surgery.

Load a test paper into QuestionCropper and mark three questions — including one crossing a page break — to see the difference directly.