H-1B Denials Are Rare — Here's What the Data Actually Shows

If you have ever searched for the "H-1B denial rate," you have probably run into two very different numbers that seem to contradict each other. One source says denials are almost unheard of. Another quotes double-digit rejection percentages. Both can be right — because they are measuring completely different stages of the process.

This piece explains what the Labor Condition Application (LCA) data actually shows about denials, why the number is so low, and — just as important — what that number does not tell you about your chances of getting an H-1B visa.

Two gates, not one

An H-1B hire clears two separate government reviews.

First, the employer files an LCA with the Department of Labor. This is a wage-and-working-conditions attestation: the employer promises to pay at least the prevailing wage for the role and location, and to meet other labor protections. The DOL reviews the form largely for completeness and internal consistency, then either certifies or denies it.

Second — and only after the LCA is certified — the employer files an actual H-1B petition with U.S. Citizenship and Immigration Services (USCIS). That is where officers scrutinize whether the job is a genuine "specialty occupation," whether the worker is qualified, and whether the employer-employee relationship holds up.

The disclosure data on this site comes from the first gate. So when we talk about a "denial rate" here, we mean LCA certification denials — not USCIS petition denials. Those are a separate process, adjudicated by a different agency, and they are not visible in this dataset. Keep that distinction in your head for the rest of this article; nearly every misunderstanding about H-1B "odds" comes from collapsing the two.

The LCA denial rate is near zero

Here is the headline: at the LCA stage, denials are rare to the point of being a rounding error for most established employers.

Look at the certification behavior of large, well-known sponsors. Among high-volume filers, the share of applications marked denied clusters far below one percent:

Two of these — Cognizant and Microsoft — show essentially no denials at all across their filings. The rest sit under roughly six-tenths of one percent. These are not cherry-picked outliers; they are among the most active employers in the program, and the pattern holds broadly across the biggest sponsors you can browse in our top H-1B sponsors rankings.

Why so low? Because the LCA is a structured attestation, not a merits test. Employers (and the immigration attorneys who file for them) know exactly what the form requires. The prevailing wage figure is looked up from official sources. The worksite and job details are filled in against a known template. A form that is complete and internally consistent gets certified. There is no adjudicator weighing whether the applicant "deserves" the job at this stage — that judgment happens later, at USCIS.

So a near-zero denial rate is less a sign of a candidate's strength and more a sign that experienced employers rarely submit malformed paperwork.

Denied is not the same as withdrawn

A quick but essential clarification: a denied LCA and a certified-withdrawn LCA are two different things, and conflating them will distort any "approval odds" you try to calculate.

A denial means the DOL declined to certify the application. A certified-withdrawn case was certified and then pulled back by the employer — often because the candidate declined, the role was cancelled, a duplicate was filed, or plans changed. Withdrawn does not mean rejected. We unpacked that nuance in detail in The Certified-Withdrawn Mystery; if you are trying to judge an employer's track record, read the withdrawn share separately from the denial share rather than lumping them together.

Reframe: it's conversion quality, not survival

If denials basically never happen at this stage, then the LCA certification rate is close to useless as a measure of "will I get in." It survives the process almost by default.

The more useful way to read this data is as conversion quality: what an employer's filings tell you about the offer itself, not about the odds of clearing a bureaucratic step. That reframes the entire question a job-seeker should be asking.

Instead of "What's this sponsor's approval rate?" — a number that is nearly always ~100% at the LCA stage — ask:

• How does the offered wage compare to the prevailing wage for this role and location? A certification only guarantees the employer attested to paying at or above prevailing wage. It says nothing about how generous the offer is above that floor.
• Which DOL wage level is the job filed at? A Level I ("entry") filing and a Level IV ("senior") filing for the same title imply very different pay and expectations.
• Is the sponsor a high-volume filer or a selective one? Both models exist, and they carry different implications for career trajectory.

That last question is worth a chart. Filing volume varies enormously across the program:

The point is not that a big filer is better or worse than a small one — it is that "denial rate" tells you nothing here, while volume, wage level, and offered wage tell you a great deal.

What the offered wage does — and doesn't — mean

Because the LCA is fundamentally a wage document, it is tempting to treat the salary on it as your paycheck. It isn't. The figure on an LCA is the offered wage the employer attests to — a gross annual (or hourly) rate, before taxes, and before the realities of bonuses, equity, cost of living, or benefits that vary by employer. Offered wage is a floor and a signal, not take-home pay.

Where the offered-wage data genuinely helps is in comparison. You can line up what an employer files for your occupation against what the occupation pays program-wide. For a role like software developers — the single largest H-1B occupation — the prevailing-wage ladder makes the entry-to-senior spread concrete:

Reading an offer against that ladder tells you something a denial rate never could: whether the job is filed as junior or senior work, and how the money compares to the market for the same code.

A note on the current year

One more guardrail before you go compute anything: the most recent fiscal year in this data (FY2026) is partial — only about two quarters have been reported so far. Denial shares, filing counts, and wage medians for a partial year can shift as the rest of the data lands. When you compare years, weight complete years more heavily, and treat the newest one as provisional.

The takeaway

• The LCA denial rate is near zero for most established sponsors — often literally 0.00%, and well under ~0.6% even for the ones that do register denials.
• That low number reflects the nature of the review (a structured wage attestation), not your visa odds. USCIS petition denials are a separate process and are not in this data.
• Withdrawn is not denied, and a partial current fiscal year should be read with caution.
• Stop treating certification rate as "approval odds." Use this data for what it is actually good at: comparing offered wages, wage levels, and filing behavior across sponsors and occupations — the real signals of offer quality.

If you want to put that into practice, start from the highest-paying H-1B jobs and the top H-1B occupations, then drill into the specific sponsor and role in your offer. That comparison will teach you far more than any denial statistic.

h1b.report is an independent project and is not affiliated with the U.S. Department of Labor or USCIS.