What Form 5500 Filings Tell You About a Company You Want to Buy

Most of the public records a searcher leans on are static or dated. A state registry tells you when a company was formed, once. An SBA loan record tells you a bank underwrote the borrower years ago. A PPP filing encodes payroll frozen in 2020 and 2021.

There is one public record that refreshes every year: the Form 5500. Any company that sponsors an employee benefit plan (a 401(k), a health plan) must file it annually with the Department of Labor, and the filings are public. For an acquisition target that has never published a number in its life, a 5500 is the closest thing to a recurring, government-filed statement that the business is still staffed and still running.

This guide covers what a 5500 actually contains, how to read it as an acquisition researcher, where the data lives, and what it honestly cannot tell you.

What a Form 5500 is

The Form 5500 series is how employee benefit plans satisfy their annual reporting duties under ERISA and the tax code. If a company sponsors a 401(k) or a similar plan, the plan files electronically through EFAST2, the system that takes these filings on behalf of the Labor Department, the IRS and the PBGC. The return is generally due on the last day of the seventh month after the plan year ends. Smaller plans, generally those with fewer than 100 participants that meet the other eligibility conditions, file the short-form 5500-SF. Larger plans file the full 5500 with its schedules.

DOL publishes the raw filings in bulk, one machine-readable file per form and plan year. The 2023 full-form file is a 29 MB zip, with matching files for the 5500-SF and for earlier years. If you would rather look up one company, the Form 5500 Series Search on EFAST2 is free and needs no login.

For a buyer researching an off-market target, the fields that matter:

Sponsor name and EIN. The company behind the plan, matchable to your target list.
Active participant count. Employees eligible for and enrolled in the plan. This is the headline number.
Plan name and type. What benefit the company offers (401(k), health, welfare).
Plan year. Which year the count describes, so every figure carries a date.

Why the participant count is the signal

A few properties make the active participant count unusually useful.

It approximates current headcount, as a floor. Participants are employees who are eligible and enrolled. Some employees never join the plan, and recent hires may not be eligible yet, so the count sits below true headcount. A plan showing 34 active participants means the company employs at least around that many people. Read it as "no smaller than" and you will never overclaim.

It refreshes annually. No other small-business public record does this. A PPP-derived payroll figure ages every year. A 5500 filed for the 2023 plan year describes the company in 2023. Two or three consecutive filings also give you a trend. Participant counts holding steady or growing across years is evidence of a stable operation, and a sharp drop is a question worth asking in diligence.

Sponsoring a plan is itself a maturity signal. A company that runs a 401(k) has payroll systems, an administrator, compliance obligations and enough organizational surplus to offer benefits. That is a different animal from a two-person shop. It's often exactly the shape of business a self-funded searcher wants: established enough to run without the owner doing everything, small enough to buy.

The honest limits

The same discipline applies here as to every public-record signal:

Absence is neutral. Most small businesses sponsor no benefit plan and file no 5500. If your target has no filing, you have learned almost nothing. It is probably just small, or offers no 401(k). Never penalize a target for a missing 5500.
It skews toward larger targets. A two-person shop rarely sponsors a benefit plan, so the smallest operators are missing from the dataset by construction. PPP data has the opposite shape. It reached far smaller borrowers, and then it stopped in 2021. Use both.
One company can file several 5500s. A sponsor with a 401(k) and a health plan files for each, and each plan covers at most the whole workforce. When you see multiple filings for one EIN in one year, take the maximum participant count as your headcount floor. Never the sum. Adding them double-counts every employee enrolled in both plans, and this is the most common reading error in the dataset.
It says nothing about revenue or profit. A participant count sizes the workforce. If you want a revenue estimate, the defensible route is headcount times the industry's receipts-per-employee from the Census Bureau's SUSB program, clearly labeled as an estimate with named inputs. We walk through that arithmetic in estimating business value from public records. EBITDA cannot be derived from a 5500 or any other public record, and no honest method claims otherwise.

Reading a 5500 in an acquisition workflow

Where the filing fits in a public-records sourcing stack:

Build the target list from the scored signals first. Registry longevity, SBA 7(a)/504 loan history and market fragmentation. Those tell you who is durable, bank-underwritten, and in a market where the acquisition math works.
Use the 5500 to size and verify. For each shortlisted target, search the sponsor name or EIN in the Form 5500 Series Search (or in the bulk files). A filing gives you a dated, current-ish headcount floor and confirms operational maturity.
Cross-check against PPP. If the target also has a 2020 or 2021 PPP record in the SBA's PPP FOIA data, you now have two independent, differently-dated statements of scale. Rough agreement raises confidence. A large gap tells you the company changed, which is worth knowing before you reach out.
Note the diligence implications for later. If you buy a company that sponsors a plan, that plan comes with the deal: administration, compliance history, a decision about continuing or terminating it. Nothing to resolve at the sourcing stage, but a 5500 in hand means your eventual diligence checklist already has a line item.
Rank and sequence outreach. The list is not the deal. Write a thesis, rank targets against it, and open owner conversations before a listing exists.

How Scouly uses 5500 data

Form 5500 filings are one of the public records behind Scouly, and the handling follows the rules above, verbatim in the pipeline code:

Scouly ingests the DOL's 5500 and 5500-SF datasets for recent plan years (currently 2022 and 2023) and attaches the participant count to matching companies as a headcount signal, framed as "active plan participants", a floor on true headcount, with the plan year and form type named.
When one sponsor files several plans in a year, Scouly keeps the maximum participant count for that company and that year, for exactly the double-counting reason above.
The headcount signal carries zero score points. Like the PPP payroll snapshot, it is evidence. It helps you size a target, and it is deliberately excluded from scoring because it is an approximation. The score rests only on longevity, SBA history and fragmentation.
Where a revenue band is shown, it is computed from the company's best headcount evidence (latest 5500 participants, else PPP jobs reported) multiplied by industry receipts-per-employee from Census SUSB, currently the 2017 receipts vintage, CPI-adjusted, with every input named in the signal and no EBITDA figure anywhere.

Sources and formulas are documented on the data page. As with the rest of the method, nothing here requires a tool. The bulk files and efast.dol.gov are free. Scouly's job is the cross-referencing at scale.

Frequently asked questions

Are Form 5500 filings really public? Yes. ERISA requires the filings, and the Labor Department publishes them. Individual filings are searchable through the Form 5500 Series Search on EFAST2, and every plan year is also released as a bulk machine-readable file you can download whole. There is no cost and no login. Anyone researching a private company can pull the sponsor's filings by name or EIN in a few minutes.

Does the participant count equal the company's employee count? Not exactly. It is a reliable floor. Active participants are employees eligible for and enrolled in the plan. Employees who opted out or aren't yet eligible are excluded, so true headcount is usually somewhat higher. Read "38 active participants" as "at least roughly 38 employees" and you'll stay on the safe side of every estimate you build on it.

My target has no Form 5500. Is that a red flag? No. Most small businesses sponsor no benefit plan and therefore file nothing. Absence means the company is probably small or simply offers no 401(k). Treat it as neutral and lean on other records for sizing, such as the state registry, SBA loan files and PPP data. Penalizing a target for a missing filing would throw out most of the market for no reason.

Can I estimate revenue or EBITDA from a 5500? Revenue, only as a clearly labeled estimate: headcount times the industry's receipts-per-employee from Census SUSB benchmarks. EBITDA, no. No public record discloses or supports a profitability figure, and that is what diligence is for. Anyone showing you an EBITDA number derived from filings alone is guessing, whatever the label says.

What is the difference between Form 5500 and Form 5500-SF? Both are annual reports filed electronically through EFAST2. The 5500-SF is the short form, used by plans with fewer than 100 participants that also meet the other eligibility conditions in the instructions. The full 5500 is for larger plans and carries more schedules. For acquisition research the field you care about, active participant count, appears on both, so treat them the same way.

Sources

Rather see the cross-referenced picture in one place? Scouly tracks off-market operators with their 5500, PPP, SBA and registry evidence attached. Browse a market on the markets page or build your thesis, free.

By Nishkal Dachepelly, founder of Scouly. Published 2026-07-02. Updated August 2026.