

You can scrape LinkedIn event attendees by opening the event's Attendees tab and logging the profiles that match your target, and a Simular AI computer agent can do the paging for you. LinkedIn has no native export for attendee lists, so the work is either manual or automated. The safe automated path drives the real LinkedIn screen, not a browser extension.
The costly move is the common one. You find a packed LinkedIn event, install a scraper extension on your main account, and pull the whole attendee list in one pass. Two days later the account is restricted, right when the follow-up matters most. The list was never the hard part. Keeping the account that lets you act on it is.
Quick answer: To scrape LinkedIn event attendees, open the event's Attendees tab, page through the list, and log the profiles that match your target. Do it by hand for a small event, or hand the repetitive paging to a Simular AI computer agent that reads each attendee inside the real LinkedIn interface. Skip Chrome extensions, which LinkedIn scans for and can restrict.
| By hand | Simular AI agent | Extensions & scrapers | |
|---|---|---|---|
| Setup | None | Show it once, about 20 minutes | Install, connect your session |
| Speed / volume | 30 to 60 profiles/hour | Human-paced, unattended, a full event overnight | Fast, but front-loads the risk |
| Cost | Free, your time | Subscription | $30 to $100+/month |
| Account risk | Lowest | Low: no extension, human pace | Highest: fingerprinted |
| Personalization | Full | Full, per attendee | Usually generic |
| Best for | Under 100 attendees | Recurring events on autopilot | A one-off dump, risk accepted |
Yield math: At a 1,200-person summit only 45 attendees are Tier 1 VCs, under 4% of the room, and paging all 1,200 by hand at 30 to 60 profiles an hour costs 20 to 40 hours. Derived estimate.
The same three steps serve very different goals. A recruiter sourcing passive tech candidates, a VP of Sales pulling RevOps buyers, and a Tier 1 VC tracking seed-stage founders all open the same Attendees tab. Only the filter and the follow-up change.
Register for the event, read the Attendees tab, and log matches yourself. Zero ban risk, capped at your own scrolling speed.
Pros: lowest risk, full context on each attendee. Cons: slow, and a large event view stops loading near 1,000 profiles, so a 3,000-person summit needs tighter filtering.
Instead of installing anything, you show a Simular agent the read-and-log routine once, and it repeats it across the attendee list for you.
Because the agent operates the same Attendees screen a person clicks, there is no browser extension in the 6,000-plus set LinkedIn fingerprints, and no third-party API to hit a cap. It reads attendees at a measured rhythm rather than blasting the list, and its guardrails pause on anything sensitive so you approve the target before it runs at volume. You get a clean, matched attendee list without a scraper signature on your account. See how the team frames this on the Simular about page.
Pros: matched list, no extension footprint, unattended. Cons: a subscription, plus about 20 minutes to teach it your routine.
Tools such as Apollo, Evaboot, PhantomBuster, Waalaxy, LinkedHelper, Dux-Soup, and Wiza can pull an attendee list from your logged-in session. Quick, but they raise the signature LinkedIn watches for.
Pros: fast, and many bundle enrichment. Cons: LinkedIn scans for prohibited software and extensions, and practitioners on Indie Hackers report accounts restricted within about 48 hours of aggressive scraping.
LinkedIn does not need to see your tool to catch it. An extension injects code into the page, which changes the DOM and the request timing in ways a normal session never does. Session-cookie and headless scrapers leave their own tells: the navigator.webdriver flag, canvas and WebGL fingerprints, and inhuman scroll cadence. That timing signature is the trigger, not the word automation. LinkedIn treats it as prohibited automated activity and can restrict the account.
What practitioners actually do:
A concrete case makes the stakes clear. A seed-stage founder registers for a 1,200-person SaaS summit. Maybe 45 attendees are Tier 1 VCs. The other 1,155 are noise for that goal, so the win is a tight filter, not a bigger dump. The mistakes that cost people their accounts are consistent: scraping the full list with an extension, moving faster than a human could, and ignoring an early restriction banner. Practitioners on Indie Hackers describe the same loop, pull about 100 leads, lose the account within 48 hours. The line worth pinning up: the attendee list is not the asset, the account that can message it is.
An event list is a different signal than a poll's voters or a group's roster. If your targets gather elsewhere, see how to extract LinkedIn poll voters or find LinkedIn group members. Once your list is built, feed it into a targeted LinkedIn prospect list.
The winning pattern: register, filter to the attendees who fit, and let a Simular Pro agent page the list on its own VM so you spend your time on the conversations, not the scrolling.
Register for the event, open its Attendees tab, and log the profiles that match your target one page at a time. LinkedIn shows attendees only to people who have registered, so sign up first.
Automated scraping of attendee data can violate LinkedIn's User Agreement, which prohibits bots and unauthorized automation. Viewing and manually noting attendees is normal use. The risk rises with extensions and headless scrapers, which LinkedIn flags as automated activity and can restrict. Stay human-paced and skip prohibited software.
Define the target before you page the list, then log only matching attendees. A recruiter keeps passive tech candidates, a VP of Sales keeps RevOps buyers, and a VC keeps seed-stage founders.
A Simular agent applies the same rule per profile and skips the rest.
You most likely used a fingerprinted extension or moved faster than a human could. LinkedIn scans for known scraping extensions and checks signals like navigator.webdriver and click cadence. Practitioners on Indie Hackers report restrictions within about 48 hours of aggressive scraping. Slow down and drop the extension.
Standardize one target definition and reuse it across every event, then let an agent do the repeat paging. Manual paging does not scale past a few events a week.
A Simular Pro agent runs the same routine across events on its cloud VM while you are offline.