H-1B approval rates explained

What the Data Hub reports

USCIS's public H-1B Employer Data Hub reports, for each employer and fiscal year, a count of approved and denied H-1B petitions, split into two categories: initial and continuing. An initial petition is the first H-1B adjudication for a given worker at that employer — a new hire, a worker moving from another status into H-1B, or a worker changing employers under the process described in the guide to changing employers. A continuing petition covers a worker already in H-1B status at that employer: an extension of stay, an amendment, or a change tied to the same underlying job. VisaBench computes an approval rate for each category separately — initial approvals divided by initial approvals plus initial denials, and the same formula run on the continuing columns — exactly as stated on the methodology page. Reading the two rates together says more than either alone: a low initial rate paired with a high continuing rate, for example, describes an employer whose extensions of existing workers clear review far more often than its petitions for new ones.

Why an approval rate is not a hiring rate

A petition is not a person and not a job offer. One employer can file several petitions tied to the same worker over a period of years — an initial petition, then one or more continuing petitions for extensions — and each is counted separately in the Data Hub. An approval also does not mean the worker ever started the job: a petition can be approved and the position still go unfilled if the worker takes another offer, does not obtain a visa stamp, or the job changes before the start date. And a denial does not necessarily mean the case ends there; a denied petition can be refiled, amended and resubmitted, or appealed, and a later filing tied to the same worker shows up as its own row in a later fiscal year. Approval rates measure how petitions filed by an employer fared in adjudication, not how many people that employer ultimately employed.

Why the rate is sometimes not published

An approval rate is undefined when an employer filed zero petitions of a given type in a fiscal year — there is no ratio to compute. VisaBench states the rate as not published rather than as zero in that case, because zero would misleadingly suggest every petition was denied. The same caution applies to reading small-volume employers generally: a single denial at an employer with only a handful of petitions swings the rate far more than the same denial would at a high-volume filer, so year-over-year rates should be read alongside the underlying counts, not as a percentage in isolation.

Why approval rates move over time

An employer's rate in a given fiscal year reflects the mix of cases it filed that year, USCIS's adjudication standards and evidentiary requirements at the time, and the specific occupations and worksites involved — factors that are not stable from year to year and are not fully visible in the aggregate count. USCIS periodically updates internal guidance on how it evaluates specialty-occupation and other H-1B eligibility questions, DOL periodically updates prevailing-wage methodology that feeds into the LCA each petition relies on, and any given employer's petitions in a year may skew toward occupations or worksites that draw more or less scrutiny than in prior years. Because so many of these inputs shift independently, a change in an employer's approval rate from one fiscal year to the next describes an outcome, not a cause — the Data Hub does not report why a particular petition was denied.

How to read one employer's trend

The most useful comparison is usually an employer against its own history: whether its initial approval rate has been broadly stable, rising, or falling across the fiscal years on record, and whether continuing approvals — extensions of workers it already sponsors — track the same pattern. A large filer's multi-year trend is more informative than any single year, because it averages out the effect of a handful of unusual cases. VisaBench's employer pages, such as Amazon, Google and Infosys, show each employer's initial and continuing rates by fiscal year side by side, drawn directly from the Data Hub, so a reader can see the shape of the trend rather than a single snapshot. The top sponsors ranking is a useful starting point for comparing rates across employers of similar filing volume, since volume itself affects how much weight a single year's rate should carry.

What an approval rate does not show

An approval rate says nothing about wages, worksite, or what happens after H-1B status — whether an employer goes on to sponsor a green card, for instance, is a separate measure covered in the guide to how the H-1B process works and in VisaBench's PERM follow-through figure. It also says nothing about the beneficiary's individual qualifications relative to another candidate at the same employer; adjudication is case-by-case, and the aggregate rate is a summary of many independent decisions, not a prediction for any one petition.

This guide is general information, not legal advice; consult a licensed immigration attorney about your case. Anyone relying on approval-rate data for a specific filing decision should also review the underlying counts and fiscal-year periods on the relevant employer page, since a rate computed from a small number of petitions can change sharply from one year to the next.