The H-1B Lottery Stops Being Random This Cycle
Wage level decides selection odds now, not chance. What sponsors need in the registration before the window opens, and why entry-level roles lose ground.
| By | Sponsorship Wire Desk — Staff Writer |
|---|---|
| Filed | 7 September 2026 |
| Read | 3 MIN |
Ask a mobility team what the H-1B cap lottery has always rewarded, and the honest answer is luck. Not this cycle. A finalized rule takes effect February 27, 2026, in time for the cap season that opens the following month. Selection no longer runs on a single random draw across every registration.
Instead, each registration earns a number of chances tied to the wage level entered for the role. Under the new matrix, a job offered at the top wage tier draws four chances in the pool, the next tier down draws three, the tier below that draws two, and the bottom tier draws one. A registrant can still win at any wage level. The odds simply stop being equal.
The logic behind the change is straightforward, even if the effect on any single sponsor is not: regulators are using pay as a stand-in for skill, on the theory that employers will stretch their offers for candidates they genuinely cannot do without. Whether that assumption holds across every occupation is a fair question, and not one this rule pretends to answer. What is not up for debate is the mechanics on the ground, which every sponsor now has to plan around.
None of this displaces the beneficiary-centric registration model adopted in earlier cycles. A candidate still gets exactly one shot at the lottery no matter how many employers file on their behalf, and the wage weighting sits on top of that rule rather than replacing it. Register the same person through two different job offers at two different wage levels, and the system does not add the chances together: the maximum weight any beneficiary can receive is four entries for a Level IV wage, and a lower-paying backup offer pulls that ceiling down rather than adding to it. Candidates fielding more than one offer, and the recruiters courting them, need to understand that a modest backup role can quietly drag down the odds attached to the job the candidate actually wants.
The mechanics sponsors need to get right
The wage level a sponsor enters is not a formality anymore; it is the entire selection engine. That number comes from the Department of Labor's occupational wage data for the role's classification and location, so the SOC code chosen at registration now carries more weight than it ever has. Pick a code that understates the job's seniority and the registration draws fewer chances than the role actually supports. Pick one that overstates it, and a mismatch between the registration and the eventual petition becomes its own liability: officials may deny or revoke a filing when the petition submitted after selection does not line up with what was entered at registration, if the change looks designed to inflate the odds after the fact. The registration and the petition need to tell the same story about the job, from day one.
Regulators also tightened the calendar around the process itself. Registration dates now have to be published well ahead of time, a fixed lead period measured in weeks, not the short notice sponsors sometimes got in past cycles. That is a small mercy for HR teams that have to clear budget and paperwork before the window opens, but it does not change the harder problem underneath: entry-level roles are about to get squeezed.
Who loses ground, and what to do about it
The agency's own modeling expects a meaningful drop in registrations selected at the lowest wage tier. That matters most for the employers who lean on it: new-graduate hiring programs, roles tied to practical training, and any position priced at the entry level because the work genuinely is entry level. None of that disappears. It just gets harder to win. Cap-exempt sponsors, universities, nonprofit research organizations and the like, sit outside this entire mechanism and keep filing the way they always have.
The workaround is not to inflate a job's wage level to chase better odds; that is exactly the pattern regulators say they will scrutinize after selection. The real workaround is upstream of registration: benchmark the role honestly against the wage data before the window opens, decide which roles can bear a higher, defensible wage level, and stop treating the registration as a placeholder to be fixed later. By the time the cap season opens, the wage level on file should already be the wage level counsel is prepared to defend in the petition.
Sponsors who treated registration as a formality now have a harder job: proving, months in advance, that the number they entered was never a guess.