Is it legal to scrape Yelp data?

Scraping publicly visible business listings (no login) generally does not violate the US Computer Fraud and Abuse Act under hiQ v. LinkedIn (9th Cir. 2022), but it can still breach Yelp's Terms of Service and its robots.txt, which prohibit automated access except for whitelisted crawlers. For contractually clean data, the Yelp Fusion API is the sanctioned path. Stick to public business data, treat reviewer identities as personal data under GDPR and CCPA, rate-limit so you don't degrade the service, and get legal advice before any commercial use.

What's the most reliable way to extract Yelp business data?

Read the schema.org JSON-LD block Yelp embeds in each business page instead of targeting CSS classes. That application/ld+json script carries the name, rating, review count, price range, address, and phone already parsed, and because it's a web standard Yelp can't freely rename it the way it rotates its generated class names. Hours are the exception and come from the rendered hours table. This is what makes yelp business data scraping durable across Yelp's frequent front-end changes.

How do I scrape Yelp reviews, and can I get all of them?

Yelp seeds the business page JSON-LD with the first page of reviews, and you page the rest by adding ?start=N in steps of 10, sorting with sort_by. Each page re-renders its own 10 reviews, so you loop until you reach the total review count. The Yelp Fusion API only returns three truncated excerpts per business, which is exactly why people scrape the full set, but remember review text is personal data, so be deliberate about what you store.

Why does my Yelp scraper keep getting a "Press and Hold" captcha?

That's PerimeterX (HUMAN Security), Yelp's bot wall, and it triggers on datacenter IPs and obvious headless clients. Fix it with residential IPs and a real rendered browser: set premium_proxy=true and render_js=true on the SparkProxy Scraping API so requests arrive from a residential exit with a genuine browser fingerprint. Detect the challenge page in your code and retry instead of saving it as data.

How many businesses can one Yelp search return?

A single Yelp search caps at roughly 24 pages of 10 results, about 240 businesses, no matter how many exist in that area. Paging further just returns the last page again. To cover a whole metro, split the query into narrower searches by neighborhood or sub-city, run each one, and dedupe the results by business alias. This tiling approach is the standard way to beat the cap when you extract Yelp listings at scale.

Does the Yelp Fusion API make scraping unnecessary?

For a lot of use cases, yes, and it's the compliant choice. The gap is reviews: the Fusion reviews endpoint returns at most three excerpts per business, each truncated to about 160 characters, and it meters you with a daily call quota. If you need full review text, every review, or fields the API doesn't expose, scraping public pages is the fallback, subject to the legal guardrails above.

Guides

How to Scrape Yelp Data (Listings and Reviews)

Scrape Yelp data the durable way: pull business name, rating, review count, category, address, phone, and hours from the page's JSON-LD, past PerimeterX.

S SparkProxyJul 26, 2026 - 08:30 0 20 min read

How to Scrape Yelp Data (Listings and Reviews)

To scrape Yelp data without your parser rotting in a month, skip the obfuscated React class names and read the schema.org JSON-LD block Yelp ships inside every business page. That one block hands you the name, rating, review count, price range, address, and phone as clean structured data, so most of the fields people fight over are already parsed for you. This guide walks the full pipeline for public Yelp business data and reviews: which fields to pull and where they live, how to page through search results and past the roughly 240-result cap, how to get through the PerimeterX bot wall, and how to collect reviews without stepping on privacy law. Every request uses SparkProxy's Scraping API, so browser rendering and residential IP rotation are request parameters instead of infrastructure you babysit.

Is scraping Yelp data legal?

Yelp business listings are public, so anyone can open a page without logging in. That settles the access question but not the permission question, and the two are separate, so get the framing straight before you write code.

In the United States, the Ninth Circuit's decision in hiQ Labs v. LinkedIn (2022) held that scraping data that's publicly accessible, meaning no login and no authentication wall, generally does not violate the Computer Fraud and Abuse Act. That ruling is about unauthorized access, not a license to copy whatever you like. Yelp's Terms of Service separately prohibit automated collection, and its robots.txt is unusually blunt about it: the file states that use of "any robot, spider ... or other automated device, process or means to access, retrieve, copy, scrape, or index any portion of the service ... is prohibited, except as expressly permitted by Yelp," and it only whitelists a few crawlers such as Googlebot on specific paths. So scraping Yelp can breach its terms and its stated crawl policy even where it clears the CFAA bar. Different questions, different answers.

Yelp also ships a sanctioned path: the Yelp Fusion API (developers.yelp.com) returns business search, business details, and reviews under a usage agreement and a daily call quota. There's one gap that pushes people toward scraping in the first place. The Fusion reviews endpoint returns at most three review excerpts per business, and each excerpt is truncated to roughly 160 characters. If your analysis needs full review text or every review on a listing, the API simply won't give it to you, and that is the narrow, gray-area space this guide operates in for research and analysis.

Guardrails that keep a Yelp project defensible:

Collect public business data only: name, category, rating, review count, address, phone, hours, price. Nothing that sits behind a login.
Treat reviews as personal data. A review ties a real name to an opinion, which is personal data under GDPR and CCPA. If you don't need reviewer identities, don't store them, and anonymize what you keep.
Read and honor robots.txt and any Crawl-delay. Rate-limit hard and back off on errors so you never degrade the service for the people actually using it.
If the data feeds a commercial product, the Fusion API is the clean route, and you should run scraping past a lawyer. This is engineering guidance, not legal advice.

What data you can extract (fields reference)

A Yelp business page carries almost everything you'd want in a single embedded JSON-LD block, plus a couple of fields that live in the rendered HTML. The search results page is thinner: it's really a list of links to business pages, so the durable pattern is to harvest business aliases from search, then follow each one for the full record. Here's the reference set worth pulling, with the stable way to get each field as of mid-2026.

Field	Where it lives	How to get it	Notes
Business name	JSON-LD `name`; page `h1`	Parse the `application/ld+json` block	JSON-LD is the most stable source
Business alias	canonical URL `/biz/`	`link[rel="canonical"]` or the `/biz/` href	Your stable per-business identifier
Rating	JSON-LD `aggregateRating.ratingValue`	float, 1.0 to 5.0	Search cards also expose it in an `aria-label`
Review count	JSON-LD `aggregateRating.reviewCount`	integer	Matches the "N reviews" on the page
Category	JSON-LD `@type` plus category links	e.g. `Restaurant`, "Italian, Pizza"	Category anchors carry `find_desc=`
Price range	JSON-LD `priceRange`	`$`, `$$`, `$$$`, `$$$$`	Not always present
Address	JSON-LD `address` (PostalAddress)	street, locality, region, postal code	Cleanly split into fields
Phone	JSON-LD `telephone`	formatted string
Hours	rendered hours table rows	day plus open/close text	Usually NOT in JSON-LD; needs render
Reviews (sample)	JSON-LD `review[]`	first page of reviews	Full set via `?start=` pagination

Two rows are the reason this guide leans on JSON-LD. The rating, review count, address, phone, and price are handed to you already parsed inside the

How to Scrape Yelp Data (Listings and Reviews)

# Is scraping Yelp data legal?

# What data you can extract (fields reference)

Is scraping Yelp data legal?

What data you can extract (fields reference)