Salary Data by Occupation: BLS Wages by State

Salary benchmarking data for recruiters, HR teams and compensation analysts. Pick a job and get back what it pays in the U.S.: mean and median annual salary, hourly wage, and how many people work in it, nationally or in any state.

The figures are the official U.S. Bureau of Labor Statistics salary statistics from the Occupational Employment and Wage Statistics (OEWS) survey, which collects pay data from employers, not from self-reported salary sites.

What you get

One row of salary statistics per job and location:

• Mean and median annual salary
• Mean and median hourly wage
• Employment: the estimated number of people employed in that job
• The year the figures come from

Who this is for

• Recruiters checking what a role pays in a state before making an offer or advising a client.
• HR and compensation teams building salary bands and pay ranges from official government data.
• Job boards and career sites showing typical pay for a job, or comparing it across states.
• Researchers and journalists covering wages by occupation and state.

Input

Each job is identified by its SOC code (Standard Occupational Classification), the code the U.S. government uses for every occupation. Find any job's code at bls.gov/soc. Some common ones:

Job	SOC code
Software developers	15-1252
Registered nurses	29-1141
Accountants and auditors	13-2011
Human resources specialists	13-1071
General and operations managers	11-1021
Customer service representatives	43-4051
Heavy and tractor-trailer truck drivers	53-3032
Retail salespersons	41-2031

Field	Type	Description
socCode + stateCode	string	Shortcut for a single job: "15-1252", plus an optional state.
occupations	array	Several jobs in one run: [{ "socCode": "15-1252", "stateCode"?: "TX" }]. O*NET-style codes like "15-1252.00" work too, and a plain list of codes (["15-1252", "29-1141"]) is fine. Omit stateCode for national data, or give a two-letter code ("TX") or the state's name ("Texas").
apiKey	string (optional)	Free registration key from bls.gov/developers, raises the daily request limit and historical range. Not required for small batches.

{
  "occupations": [
    { "socCode": "15-1252" },
    { "socCode": "15-1252", "stateCode": "TX" }
  ]
}

Output

One record per occupation lookup:

{
  "socCode": "15-1252",
  "stateCode": null,
  "found": true,
  "annualMeanWage": 148100,
  "employment": 1687890,
  "hourlyMeanWage": 71.2,
  "annualMedianWage": 135980,
  "hourlyMedianWage": 65.38,
  "dataYear": "2025",
  "checkedAt": "2026-09-19T12:00:00.000Z"
}

Fields come back as null if that specific measure has no data available for the requested occupation/area combination (common for niche occupations or small states), found is false if none of the requested measures returned data at all.

An entry that can't be looked up doesn't stop the run. It comes back with found: false and an error that says what to fix, for example a job title instead of a SOC code, or a state that doesn't exist. The other lookups in the same run carry on as normal.

How it works

Direct calls to the official BLS Public Data API's OEWS series, no scraping, no proxy. Each request builds the OEWS series ID for employment, annual mean wage, annual median wage, hourly mean wage, and hourly median wage, and fetches them in batched calls of up to 5 occupation/area lookups each (10 with a registration key).

Rate limits: unregistered use is capped at 25 series requests per day and 10 years of history. A free registration key (see Input above) raises that to 500 requests per day with a longer history window.

Pricing note

Billed per occupation lookup, not per wage field returned: one charge per socCode/stateCode pair that BLS answers, whether all five wage measures come back populated or several are null for that occupation/area. Entries with an error (input that isn't a valid code or state, or a lookup BLS refused, for example once its daily limit is used up) are never charged.

Related products

• US Labor Data for broader BLS labor market indicators, not tied to one job's pay