How to Read Salary Percentiles (and Why the Average Is Misleading)

Type almost any job title into a search engine and you will get a single number: "the average software developer earns $132,000," "the average nurse earns $94,000." That number is easy to quote and almost useless on its own. It hides the difference between a first-year hire in Mississippi and a twenty-year veteran in San Francisco, between a small nonprofit and a Fortune 500 employer. If you make career or negotiation decisions based on a single average, you are navigating with half a map.

The fix is to read the full wage distribution — the percentiles. The Bureau of Labor Statistics publishes five points on that distribution for nearly every occupation it tracks, and once you know how to read them, you can answer much sharper questions: What does entry-level actually pay here? What is a realistic ceiling? Is this offer genuinely good, or just not bad?

What each percentile actually means

A percentile is a rank, not a grade. The 25th percentile wage is the amount that 25% of workers in an occupation earn less than and 75% earn more than. Concretely, the five published points break down like this:

• 10th percentile (p10). Only one in ten workers earns less than this. Think of it as the floor for the occupation: trainees, brand-new hires, part-year workers, and jobs in the lowest-paying regions and industries cluster here.
• 25th percentile (p25). The lower-middle of the field. A realistic benchmark for someone with limited experience or working in a lower-cost area.
• 50th percentile — the median. Half earn more, half earn less. This is the single most honest answer to "what does this job pay?"
• 75th percentile (p75). The upper-middle. Experienced workers, those in high-paying industries, or those in expensive metro areas typically sit here.
• 90th percentile (p90). Only one in ten earns more. This is a realistic long-run ceiling for most people in the occupation — not the absolute maximum, but the point beyond which you are usually looking at management, rare specialization, or unusually generous employers.

The distance between p10 and p90 is itself a piece of information: it tells you how much the occupation rewards experience, location, and industry choice. Some occupations are tightly compressed — a postal carrier at the 90th percentile earns perhaps 50% more than one at the 10th. Others are wide open: a financial advisor at p90 can earn four or five times the p10 wage. A wide spread means your decisions (where you work, which industry, how you negotiate) matter enormously. A narrow spread means the pay is largely set by the occupation itself.

Mean vs. median: why the two "averages" disagree

Most salary pages on Wage Atlas show both a mean (the arithmetic average) and a median. They are rarely equal, and the gap is informative. Wages have a hard floor — nobody earns a negative salary — but no ceiling. A handful of workers earning $400,000 in an occupation where most earn $80,000 pulls the mean up substantially while leaving the median untouched.

As a rule of thumb: when the mean sits well above the median, the occupation has a long right tail of high earners, and the median is the better description of a typical paycheck. When mean and median are close, pay is evenly distributed and either number is fine. You will almost never see the mean below the median in wage data, because the skew nearly always runs toward high earners.

This is also why "average salary" headlines mislead. The average is the number most inflated by outliers — exactly the workers whose situation probably does not resemble yours.

A worked example

Suppose you are evaluating an offer as a registered nurse, and the state-level distribution looks like this (figures rounded for clarity):

• 10th percentile: $66,000
• 25th percentile: $76,000
• Median: $89,000
• 75th percentile: $104,000
• 90th percentile: $121,000

An offer of $84,000 now has context. It sits between the 25th percentile and the median — a below-typical wage for the state. For a new graduate, that may be entirely reasonable; for a nurse with eight years of ICU experience, it is a signal to negotiate or keep looking. Without the percentiles, "$84,000" is just a number that sounds large. With them, it is a position on a map.

Notice what else the distribution tells you: the spread from p10 to p90 is under 2x, which is relatively compressed. Nursing pay is structured by employers and union scales; the biggest lever for a nurse who wants to move up the distribution is usually location andshift/specialty, not negotiation alone. Compare that with sales occupations, where the p90 can be several multiples of the p10 and individual performance dominates.

How experience, industry, and geography move you along the curve

The OEWS data that powers Wage Atlas does not break wages down by years of experience — no government survey does at this scale. But the percentile distribution is a reliable proxy, because the same forces show up in it again and again:

Experience moves most workers from around p10–p25 at entry toward p50–p75 at mid-career. Reaching p90 usually requires something beyond tenure: a leadership role, a high-paying industry, or an expensive labor market.

Industry can matter as much as the job title. The same database administrator earns very different wages in local government, in a hospital system, and at a software company. On any Wage Atlas salary page you can see the industry breakdown for an occupation — if your industry pays below the occupation's median, switching industries is often a larger raise than any internal promotion.

Geography shifts the entire curve. A p50 wage in one state can exceed a p75 wage in another. That is precisely why we publish state and metro distributions rather than a single national figure — and why comparing your salary to a national average can leave you either falsely satisfied or needlessly alarmed. Our best states tool ranks every state for a given occupation, and the cost-of-living companion article explains how to adjust those comparisons for regional prices.

Reading percentiles when you evaluate an offer

A practical sequence we recommend:

1. Find your occupation and state using the salary search. Use the state where the job is located, not where you live now.
2. Anchor on the median. This is "typical" for the occupation in that market.
3. Place yourself honestly. Entry level with no specialization? Benchmark p10–p25. Several years in with in-demand skills? p50–p75. Deep specialist or proven leader? p75–p90.
4. Check the industry mix. If the employer is in an industry that pays above the occupation median, expect the offer to be above median too — and treat a median offer as low.
5. Adjust for cost of living if you are comparing offers across states. The RPP toggle on our salary pages does this in one click.

One caution: percentile data describes all current workers, including people hired years ago at wages that have since grown. In fast-moving labor markets, new-hire offers can run ahead of the published distribution. Treat OEWS percentiles as a reliable floor and structure, then layer on current job-posting data for the final read.

The bottom line

A single average flattens an occupation into one number and erases everything that makes your situation specific. The five percentiles restore that information: where the floor is, where the ceiling is, what typical looks like, and how much room the occupation gives you to climb. Every salary page on Wage Atlas shows the full distribution for this reason. Learn to read it, and you will never again have to ask whether a number is "good" — you will be able to see exactly where it lands.