Income Percentile Calculator

Introduction to income percentile ranking in the United States

An income percentile answers one narrow, factual question: out of every hundred earning units in a defined population, how many take home less than you do? If your income sits at the 78th percentile, then 78 of every 100 units in that population report less income and 22 report more. The number is descriptive, not normative — it says nothing about whether the income is adequate, fair, or comfortable where you live.

The reason percentile calculators disagree so violently with one another is almost never the arithmetic. It is the universe: the population being ranked. Ranking a salary against every person aged 15 and over who reported any income at all — a group that includes part-year workers, students with summer jobs, and retirees living on Social Security — produces a far flattering answer than ranking the same salary against full-time year-round workers, and a completely different answer again from ranking it against whole households, many of which contain two earners. This calculator makes the universe an explicit input rather than a hidden assumption, and it carries three separate published distributions instead of a single smoothed curve.

All three distributions come from the same survey: the U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement (CPS ASEC), fielded in March 2025 and covering income year 2024. The income concept is Census money income: gross, pre-tax cash income from all sources over the full calendar year.

How to use the calculator without mismatching your income to the wrong universe

1. Pick the comparison universe first. If you are a salaried employee who worked all year, choose full-time, year-round workers — that is the closest apples-to-apples group. If you want the headline figure most often quoted in news articles about "individual income," choose all individuals age 15+ with income. If you are comparing a whole family's budget, choose households and enter the combined income of everyone living at the address.
2. Enter gross annual income in whole dollars. Use the pre-tax figure for the full calendar year, not a monthly figure, not an hourly rate, and not take-home pay. For a household, add every resident's money income together.
3. Read the percentile alongside the headcount. The result panel reports how many millions of people or households fall above and below you, which is often more intuitive than the percentile itself.
4. Check the cross-universe table. The same income is scored against all three distributions at once, so you can see immediately how much of your rank is an artefact of the comparison group.
5. Use the milestone table to plan. It shows the income required to reach the 25th, 50th, 75th, 90th, 95th and 99th percentiles of the selected universe, and how far your current income is from each.

The income box accepts decimals and recalculates as you type once a first result exists, so you can treat it as a what-if slider. Every calculation happens in your browser; nothing is transmitted.

The percentile formula: bracket interpolation plus a Pareto upper tail

Census does not publish a percentile for every dollar amount. It publishes a grouped frequency table: the number of people or households whose income falls inside each of roughly forty brackets. For individuals the brackets run in 2,500-dollar steps from 1 dollar to 99,999 dollars, then widen to 100,000–149,999, 150,000–199,999, 200,000–249,999, and an open-ended 250,000 dollars and over. For households the brackets run in 5,000-dollar steps to 199,999 dollars, then 200,000–249,999 and 250,000 and over.

The first step is to convert those counts into a cumulative distribution. Let n_i be the count in bracket i, b_i the lower bound of that bracket, and N the total count in the universe. The cumulative percentage at the upper bound of bracket i is

Inside a bracket the calculator assumes a locally uniform density, which is the conventional treatment when bins are narrow relative to the spread of the data. For an income x lying between bounds b_i and b_i+1, the percentile is the linear interpolation

That handles every income up to the start of the open-ended top category. Above it, linear interpolation is impossible because there is no upper bound to interpolate towards — and this is exactly where naive calculators fail, either flattening the tail into a straight line or clamping the answer at a nonsensical 100th percentile. The standard remedy, used by statistical agencies and in the inequality literature, is to close the tail with a Pareto distribution. Writing x_m for the lower bound of the open bracket and C_m for the cumulative percentage reached at that point, the survival function gives

The shape parameter α is not guessed. Census publishes the mean income of the open-ended bracket, and for a Pareto variable the mean above a threshold is α x_m / (α − 1). Inverting that identity pins the exponent to the published mean μ_m:

For individuals the published mean of the 250,000-and-over category is about 446,600 dollars, giving α = 2.27. For households the 250,000-and-over mean of 421,600 dollars gives α = 2.46, and for full-time year-round workers the implied mean of the 100,000-and-over category gives α = 2.09. A lower exponent means a fatter tail, which is why individual and worker incomes thin out more slowly at the top than household incomes do relative to their own threshold.

The same machinery run backwards answers the planning question "what income would put me at percentile p?". Below the tail threshold the linear step is simply inverted; above it,

Finally, the percentile is turned into a headcount, which most people find easier to interpret than a rank. With N units in the universe, the number earning more than you is

Worked example: ranking a 120,000 dollar individual income

Suppose an individual reported 120,000 dollars of money income in 2024 and wants a rank against all people aged 15 and over with income. The relevant slice of the reconstructed cumulative table is:

1. The income falls in the 100,000–149,999 bracket, so b_i = 100,000, b_i+1 = 150,000, C_i = 81.17 and C_i+1 = 90.87.
2. Position within the bracket: (120,000 − 100,000) ÷ (150,000 − 100,000) = 20,000 ÷ 50,000 = 0.400.
3. Percentile gap spanned by the bracket: 90.87 − 81.17 = 9.70 points.
4. Interpolated percentile: 81.17 + 0.400 × 9.70 = 81.17 + 3.88 = 85.05.
5. Headcount above: (100 − 85.05) ÷ 100 × 246.5 million = 36.9 million individuals with income out-earn this person.

Now push the same person to 600,000 dollars, which lies above the 250,000-dollar open bracket. Here the Pareto branch applies: 100 − (100 − 97.15) × (250,000 ÷ 600,000)^2.27 = 100 − 2.85 × 0.137 = 99.61. Roughly 960,000 individuals still report more. A model that clamped the answer to 100.0 would have told this earner that literally nobody in the country out-earns them, which is false by six orders of magnitude.

Interpreting the number: what a percentile does and does not tell you

A percentile is a position, not a standard of living. Three cautions matter most:

• The distribution is extremely right-skewed. Moving from the 50th to the 60th percentile of individual income costs about 11,000 dollars; moving from the 90th to the 95th costs about 52,000 dollars; moving from the 99th to the 99.9th costs roughly 700,000 dollars. Equal percentile steps are not equal dollar steps.
• Geography is missing. These are national distributions. The same 110,000 dollars ranks very differently against the local distribution in a coastal metropolitan area than in a rural county, and cost of living is not adjusted at all.
• Household size is missing. A household percentile treats a single person and a family of five identically. Equivalence-scaled measures such as Census supplemental poverty statistics or OECD-modified scales exist precisely because raw household income overstates the living standard of large households.

As a rough interpretive guide for the selected universe: below the 25th percentile usually reflects part-year work, part-time hours or reliance on transfer income; the 25th to 50th band is the lower-middle of the distribution; the 50th to 75th band typically corresponds to established full-time employment; the 75th to 90th band is professional and managerial territory; and above the 95th percentile small dollar differences translate into large rank differences because the density thins out so quickly.

How the three published universes compare

The table below is the single most useful summary on this page. Every figure is either published directly by Census or reconstructed from the published bracket counts using the formulas above.

The medians shown for individuals and for full-time year-round workers are the values Census prints in PINC-01; the household median is the value printed in the P60-286 report. Reconstructing those same medians from the bracket counts returns $45,145, $67,450 and $83,988 respectively, which is the accuracy check that matters most.

Validating the model against Census-published breakpoints

Interpolation is only trustworthy if it reproduces the breakpoints the agency publishes independently. Census Table H-1 gives household income limits for each fifth of households without using any bracket table at all, so it is a genuine out-of-sample test of the method used here.

Agreement within one percent across the whole range, including the Pareto-modelled 95th percentile, is about as good as grouped-data interpolation gets. The residual differences are smaller than the CPS sampling error on the published limits themselves.

Assumptions and limitations you should weigh before quoting the result

• Uniform density inside brackets. Real income density falls across most brackets rather than staying flat, which biases interpolated percentiles very slightly. With 2,500-dollar bins the effect is a small fraction of a percentile point; it grows in the 50,000-dollar-wide brackets above 100,000 dollars.
• Pareto shape above the top bracket. The upper tail is genuinely close to Pareto in tax-record studies, but CPS survey data under-report very high incomes, so the tail exponent estimated from survey means is approximate. Treat any result above the 99th percentile as indicative only.
• Top-coding and disclosure protection. Census applies rank-proximity swapping to very high reported incomes to protect respondent confidentiality. This deliberately blurs individual records at the top of the distribution.
• Sampling and non-sampling error. The CPS ASEC is a sample of roughly 90,000 addresses. Published medians carry standard errors of a few hundred dollars, and income from self-employment, interest and transfers is systematically under-reported in household surveys relative to administrative totals.
• Money income excludes a lot. Capital gains, employer-paid health insurance, the value of food and housing assistance, and refundable tax credits are all outside the concept. Anyone with substantial capital gains will rank lower here than a tax-based measure would place them.
• One year, one country, nominal dollars. The figures are 2024 U.S. dollars for the 2024 calendar year. They are not inflation-adjusted to today and do not apply outside the United States. Comparing your own percentile across years mixes your income growth with distributional change.
• No demographic conditioning. Rank is computed against the whole universe, not against your age band, occupation, education or state. A 28-year-old at the 60th percentile overall is in a very different position from a 58-year-old at the same rank.

Sources. U.S. Census Bureau, PINC-01: Selected Characteristics of People 15 Years Old and Over by Total Money Income in 2024 and PINC-11: Income Distribution to $250,000 or More for Males and Females: 2024, Current Population Survey 2025 Annual Social and Economic Supplement (PINC-01, PINC-11). U.S. Census Bureau, HINC-06: Income Distribution to $250,000 or More for Households: 2024 (HINC-06). U.S. Census Bureau, Historical Income Tables: Households, Table H-1, Income Limits for Each Fifth and Top 5 Percent of All Households: 1967 to 2024 (Historical Income Tables). U.S. Census Bureau, Guzman and Kollar, Income in the United States: 2024, Current Population Reports P60-286, September 2025 (P60-286). Definitions of money income and of the full-time year-round universe follow the CPS ASEC technical documentation.