H-1B lottery odds

Estimate your chance of being selected in the H-1B cap lottery, by wage level and degree, under the current weighted selection rule.

Your situation

Wage level of the offered salary

The level on the registration is not a free choice: it must be the highest OEWS wage level that your offered salary equals or exceeds for the job's SOC code in the area of intended employment (below Level I, e.g. a non-OEWS survey wage, counts as Level I; multiple work locations use the lowest level the salary meets; a salary range uses its lowest figure). Look up the four levels for your job: DOL wage search (FLC Data Center) ↗

US master’s degree or higher?

From a US nonprofit or public university. Makes you eligible for the extra 20,000 master’s cap.

Default: FY2027 registrations (~211,600). Change it to see how odds move if more or fewer people register.

Selection method

Weighted selection applies from the FY2027 season (March 2026) onward. Switch to random to compare with the old rule.

Advanced assumptions

USCIS selects more than 85,000 because many selected petitions are never filed. Default: the latest season’s total, split 65 : 20 between the two rounds. (90,740 + 27,920)

%
Registrants by wage level (%)

Your estimated chance

63.6%FY2028 · Level II · US master’s
Regular cap round
41.9%
Master’s cap round (if not picked in round 1)
37.3%

Average registrant: 56.1%

Chance of at least one selection

1 season63.6%
2 seasons86.7%
3 seasons95.2%

If you register every year with the same profile and the same competition. Useful for STEM OPT, which gives up to 3 seasons.

All wage levels

Wage levelEntriesNo US master’sUS master’sShare of registrants
Level I123.8%39.7%29%
Level II241.9%63.6%36%
Level III355.8%78.0%22%
Level IV466.3%86.7%13%

Past seasons (USCIS data)

Fiscal yearEligible registrationsUnique beneficiariesSelectedSelection rate
FY2027~211,600———
FY2026343,981~339,000118,66035.0%
FY2025470,342~442,000127,62428.9%
FY2024758,994~446,000188,40024.8%
FY2023474,421—127,60026.9%
FY2022301,447—131,92443.8%
FY2021269,424—124,41546.2%

Selected ÷ unique beneficiaries (or ÷ registrations before FY2025, when multiple registrations each had their own chance).

First season with weighted selection. USCIS has not posted its table yet; ~211,600 registrations is from a USCIS social-media post relayed by Fragomen. 66.1% of registrants had a US master’s.

How this is calculated

  1. One chance per person. Since FY2025 USCIS selects by unique beneficiary (passport), so several employers registering the same person don’t add chances.
  2. Wage weighting. From the FY2027 season (final rule Dec 29, 2025, effective Feb 27, 2026) each beneficiary is entered once per OES wage level: Level I = 1 entry, II = 2, III = 3, IV = 4. Multiple work sites use the lowest level; multiple registrations for one person use the lowest level. It is a permanent regulation, so it applies to the March 2027 (FY2028) registration too; no court has blocked it.
  3. Two rounds. Round 1 draws from everyone, master’s included, for the regular cap. Round 2 draws from the US master’s holders not picked in round 1 for the 20,000 master’s cap. Both rounds are weighted.
  4. How many are selected. USCIS over-selects to make up for petitions that are never filed. The default is the latest published count (FY2026: 118,660 beneficiaries), split 65,000 : 20,000 between the rounds: 90,740 + 27,920. The FY2027 count isn’t published yet, which is the biggest uncertainty here.
  5. Formula. Drawing weighted entries without replacement until K people are picked gives each person of weight w a chance of about π = 1 − e−λw, with λ set so the expected number picked equals K (Rosén’s asymptotic formula; the unit tests check it against an exact Monte-Carlo draw to within 2 points). Without weighting this is exactly K/N. Master’s overall = π₁ + (1 − π₁)·π₂.
  6. Who registers. Default 66.1% with a US master’s (FY2027, per the Sept 2026 proclamation). The wage-level mix 29/36/22/13% is fitted so this model reproduces the FY2027 selection profile (Level I = 17.8% of selections, Levels III+IV = 46.3%). DHS’s projection in the rule was 28/55/12/5%.
  7. Multi-year. 1 − (1 − p)n, assuming each season is the same and independent.

λ₁ = 2.718e-1, λ₂ = 2.335e-1, master’s left for round 2 ≈ 79,888.