Published April 6, 2026 · Updated annually
Salary Percentiles Explained: What the Numbers Actually Mean
Most salary websites show you a single number — the average. But averages can be misleading. BLS salary percentiles (10th, 25th, 50th, 75th, 90th) show you the full distribution of what people actually earn, giving you a much clearer picture of where your salary falls.
What Is a Salary Percentile?
A salary percentile tells you what percentage of workers in a given role earn less than a specific amount. If your salary is at the 75th percentile, it means you earn more than 75% of people in the same occupation — and 25% earn more than you.
The Bureau of Labor Statistics publishes five key percentile points for every occupation:
| Percentile | What It Means | Who's Typically Here |
|---|---|---|
| 10th | 90% of workers earn more | Entry-level, part-time, or low-cost-of-living areas |
| 25th | 75% of workers earn more | Early career (1-3 years), smaller companies |
| 50th (Median) | Half earn more, half earn less | Mid-career, typical market rate |
| 75th | Only 25% earn more | Senior level, high-demand specializations |
| 90th | Only 10% earn more | Directors, principals, top-market leaders |
Real Example: Software Developer Percentiles
Here is how software developer salaries break down in San Jose, CA, the highest-paying metro:
| Percentile | San Jose, CA | National Avg |
|---|---|---|
| 10th | $134,830 | — |
| 25th | $171,610 | — |
| 50th (Median) | $208,270 | $142,414 |
| 75th | $222,550 | — |
| 90th | $241,114 | — |
The gap between 10th and 90th percentile is $106,284 — showing massive variation within the same job title and city. This is why "average salary" is nearly useless for benchmarking your own pay.
Why Median Beats Average
The median (50th percentile) is almost always more useful than the mean (average) for salary data. Here is why:
- Outliers distort averages — a few executives earning $500K+ pull the mean up, making it seem like the "typical" worker earns more than they do.
- The median is the true midpoint — exactly half of workers earn more, half earn less. No distortion from outliers.
- BLS uses median as the primary measure — the Bureau of Labor Statistics reports median wages as the standard metric for exactly this reason.
How to Use Percentiles for Salary Negotiation
- Find your role and city — look up the exact percentile breakdown for your occupation in your metro area on SalaryTruth.
- Identify your target percentile — with 3-5 years of experience, you should reasonably aim for the 50th-75th percentile. With 10+ years, 75th-90th is achievable.
- Adjust for cost of living — if you are comparing offers between cities, use the COL-adjusted figures. A $130K offer in Austin may be worth more than $160K in San Francisco.
- Present the data — in negotiation, cite specific BLS percentile data rather than vague claims. "The 50th percentile for this role in this city is $X, and I believe my experience warrants the 75th percentile" is far more persuasive than "I think I deserve more."
Common Mistakes with Salary Data
- Comparing raw salaries across cities — always use COL-adjusted figures. $120K in Houston has more purchasing power than $160K in New York.
- Using self-reported data as your benchmark — Glassdoor and similar sites rely on voluntary submissions, which skew higher. BLS surveys 1.1 million establishments and is considered the gold standard.
- Ignoring the full range — knowing only the median tells you the midpoint, but the 10th-90th range shows what is actually achievable in your career. The 90th percentile represents a realistic ceiling for top performers.
- Forgetting total compensation — BLS data covers base salary only. Stock options, bonuses, benefits, and retirement contributions can add 20-50% on top, especially in tech.
Frequently Asked Questions
With 0-2 years of experience, the 25th-50th percentile is typical. With 3-7 years, aim for 50th-75th. With 10+ years or specialized skills, the 75th-90th percentile is achievable. These targets vary by occupation and location.
No. The 50th percentile (median) is the midpoint where half earn more and half earn less. The average (mean) can be pulled higher by outliers. For salary data, the median is almost always the more useful number because it is not distorted by extreme values.
The Bureau of Labor Statistics surveys over 1.1 million business establishments every year through the Occupational Employment and Wage Statistics (OEWS) program. This is payroll-verified data, not self-reported, making it the most reliable salary source available.