PatentBrief

Guide

How to read a patent.

Patents are written for lawyers. This guide teaches you how to actually read one — what to skip, where the legal weight lives, and how to find the gaps that matter commercially.

7 stepsStart to finish
10 min readPlain English
No legal backgroundrequired

Start with the abstract — but don't trust it

The abstract is a 150-word summary at the top of every patent. It's a useful orientation but it's not legally binding. It tells you roughly what the invention is about, which helps you decide whether to keep reading.

Never quote the abstract when analyzing what a patent covers or doesn't cover. The only thing that matters legally is the claims.

The abstract is the trailer. The claims are the movie.

Look at the drawings

Every utility patent includes diagrams. Spend two minutes with them before reading anything else. Patent drawings are usually more informative than the text — they show you the physical structure, the process flow, or the system architecture at a glance.

Each element in the drawings has a reference number (like 102 or 304). These same numbers appear throughout the written description. Once you spot them in the drawings, you can trace exactly which part of the text describes what.

Reference numbers are your map. Learn them in the drawings, then track them in the text.

Read the background — briefly

The Background section explains what problem existed before this invention. It describes the prior art — existing solutions — and their shortcomings. This tells you why the inventor thought their invention was worth patenting.

It's also strategic: what the inventor criticizes in the background often reveals what they'll claim as their improvement. If the background says 'prior systems were slow,' expect the claims to include something about speed.

Skip to the claims — this is what the patent actually protects

The claims are numbered paragraphs at the very end of a patent, and they are the only part with legal force. Everything else — the abstract, the drawings, the detailed description — exists to support and explain the claims.

Think of claims like a property deed. The detailed description is the realtor's marketing copy about the house. The claims define the actual lot lines.

Each claim is a single sentence (often a very long one) that defines one version of the invention. If a competitor's product doesn't match every element of at least one claim, there is no infringement — regardless of how similar it looks.

The claims are the patent. Everything else is context.

Find the independent claims — they're the broadest

An independent claim stands alone. It doesn't reference any other claim. These are always the most important claims because they define the widest possible scope of protection.

A dependent claim starts with something like 'The method of claim 1, wherein...' It adds a limitation that narrows the scope. Dependent claims matter because: (1) they may still be valid even if the independent claim is invalidated, and (2) they define specific embodiments the inventor wanted protected.

When you're assessing a patent's breadth, read the independent claims first. The fewer elements they contain, the broader — and stronger — the protection.

Independent claims = broadest scope. Count their elements. Fewer is more powerful.

Map what it does NOT cover

This is the analysis most patent sites skip — and the most commercially useful part. A patent only blocks what its claims specifically recite. If a competitor changes one claimed element, they're no longer infringing, even if the product is functionally identical.

To find the gaps: read each element of the independent claims and ask 'what if someone removed this element? What if they replaced it with something different?' The space outside the claims is the design-around space.

Also check the prosecution history (the back-and-forth between the inventor and the patent examiner, publicly available on Google Patents). Any claim element the inventor added to get the patent granted creates a restriction on how broadly the claim can be interpreted — this is called prosecution history estoppel.

What a patent doesn't cover is often more valuable to know than what it does.

Check the priority date and expiration

The priority date is the earliest filing date the patent can claim — it determines what counts as prior art against it. The expiration date tells you whether the patent is still in force.

A utility patent expires 20 years from the earliest non-provisional filing date, assuming maintenance fees were paid. You can verify maintenance fee status on the USPTO Patent Center database.

If the patent has expired, the invention is in the public domain. You can use it freely. This is how generic drugs, open-source hardware, and foundational technologies become available to everyone.

Apply it now

Read a real patent
with this framework.

Apple's slide-to-unlock patent — one of the most litigated patents in history. Walk through every step of this guide with a real example.

Open the patent