The percentile rank formula calculates what percentage of values in a dataset fall below a given score. The basic formula is PR = (B / N) × 100, where B is the count of values strictly below the score and N is the total number of values. When tied values exist, use the extended formula: PR = [(B + 0.5 × F) / N] × 100, where F is the frequency of the score. The result is a number from 0 to 100.
How to find the percentile rank of a score:
- 1 Sort all values in ascending order.
- 2 Count how many values are strictly below your score (B).
- 3 Count the total number of values (N).
- 4 Calculate: PR = (B ÷ N) × 100.
What Is Percentile Rank?
Percentile rank is a number between 0 and 100 that shows the percentage of values in a dataset that fall below a particular score. If your test score has a percentile rank of 72, you scored higher than 72% of the people who took the same test. It's a positional measure — not a score, not a grade, but a rank expressed as a percentage.
One thing regularly trips students up: percentile rank and percentile are not the same thing, even though the words sound almost identical. Here's the clearest way to tell them apart:
| Percentile | Percentile Rank | |
|---|---|---|
| What it is | A data value (a score or measurement) | A percentage position (0–100) |
| Example | "The 90th percentile score is 1400" | "Your score has a percentile rank of 90" |
| Direction | Input → tells you a score at a rank | Output → tells you the rank of a score |
Put simply, a percentile answers "What score corresponds to this position?" A percentile rank answers "What position does this score hold?" They are inverse operations of each other — a distinction explored further in the Percentile vs. Percentile Rank section below.
Percentile rank belongs to the broader family of positional statistics covered in descriptive statistics. If you're seeing it for the first time, the percentiles page gives useful background on how the entire percentile system works.
The Percentile Rank Formula
Basic Formula (No Tied Values)
When no other values in the dataset equal the score you are ranking, use the basic formula:
PR = percentile rank (0–100)
B = count of values strictly below the score
N = total number of values in the dataset
The variable names here are chosen for clarity: B for below, N for the total number of values. Some textbooks use c or L in place of B — the formula works the same regardless of the letters used.
Extended Formula (For Tied or Repeated Values)
When two or more values in the dataset equal the score, the basic formula understates where that score actually sits. The extended formula handles this by giving each tied value equal partial credit:
B = count of values strictly below the score
F = frequency — how many times the score appears
N = total number of values in the dataset
To see why the extended formula matters, here's a direct comparison using the same dataset:
| Formula | Dataset: {60, 72, 72, 72, 85, 90, 90, 95} | Percentile rank of 72 |
|---|---|---|
| Basic (B/N × 100) | B = 1, N = 8 | 12.5 |
| Extended [(B + 0.5F)/N × 100] | B = 1, F = 3, N = 8 | 31.25 |
The basic formula only counts the single value below 72, ignoring the fact that 72 itself appears three times. The extended formula places all three occurrences at the midpoint of the range they occupy — which is both more accurate and the standard method in formal testing contexts.
Percentile rank does not span a clean 0–100 range for values within a closed dataset. The minimum possible PR is 0 (the lowest value has zero values below it). The maximum is (N−1)/N × 100 — not 100. A score cannot rank above itself. For a dataset of 10 values, the highest possible percentile rank is 90, not 100.
Step-by-Step Worked Examples
Example 1 — No Tied Values (Basic Formula)
Find the percentile rank of a score of 78 in a class of 10 students.
Test scores: 55, 62, 70, 73, 75, 78, 82, 85, 88, 95
Confirm sorted order. The list is already in ascending order: 55, 62, 70, 73, 75, 78, 82, 85, 88, 95.
Count values strictly below 78. Values below 78: 55, 62, 70, 73, 75 → B = 5
Count total values. There are 10 scores → N = 10
Apply the formula. PR = (5 / 10) × 100 = 50
✓ A score of 78 is at the 50th percentile rank — higher than exactly 50% of the class.
Example 2 — With Tied Values (Extended Formula)
Find the percentile rank of 72 in a dataset with repeated values.
Scores: 60, 72, 72, 72, 85, 90, 90, 95
Sort. Already in ascending order: 60, 72, 72, 72, 85, 90, 90, 95.
Basic formula first — to show the problem. B = 1 (only 60 is strictly below 72). PR = (1/8) × 100 = 12.5. This ignores the three tied values at 72 and understates the actual standing.
Identify B, F, and N for the extended formula. B = 1, F = 3 (three 72s in the dataset), N = 8.
Apply the extended formula. PR = [(1 + 0.5 × 3) / 8] × 100 = [(1 + 1.5) / 8] × 100 = [2.5 / 8] × 100 = 31.25
✓ Using the extended formula, 72 sits at a percentile rank of 31.25. This is the method SAT and GRE score reports use to handle ties across a large test population.
Example 3 — Real-World: Test Score Report
A student scores 84 in a class of 20. One other student also scored 84.
Class of 20 students. 15 scored strictly below 84; one other student also scored 84.
Identify B, F, and N. B = 15 (scores strictly below 84), F = 2 (two students scored 84), N = 20.
Apply the extended formula (use it because F = 2, not 1).
PR = [(15 + 0.5 × 2) / 20] × 100 = [16 / 20] × 100 = 80
✓ This student's score of 84 is at the 80th percentile rank.
Working with a larger dataset? The percentile calculator handles the counting automatically.
Use the Percentile Calculator →Percentile Rank vs. Percentile — The Key Distinction
These two terms appear on the same standardized test score reports, which is where the confusion starts. When a student receives their SAT results, the report might show a scaled score of 1250 alongside a percentile rank of 81. Those are two different pieces of information.
The scaled score (1250) is a percentile value — it is a data point on the measurement scale. The percentile rank (81) is a position — it tells the student that their score of 1250 was higher than 81% of test-takers in the reference group for that year.
The directionality is what separates them. A percentile answers: "What score corresponds to this rank?" A percentile rank answers: "What rank does this score hold?" If you already have a percentile rank and want to find the corresponding data value, that's a percentile lookup — the opposite operation. The percentile calculator handles both directions.
For another way to standardize and compare scores, the z-score converts values to standard deviations above or below the mean. Z-scores and percentile ranks both express position within a distribution, just on different scales. The relationship between them runs through the standard normal table.
How to Calculate Percentile Rank in Excel
Excel has specific functions for this, and the naming is easy to mix up. The correct function for percentile rank is PERCENTRANK.INC. It takes the array of values and the score, then returns a number between 0 and 1. Multiply by 100 to get the percentage:
' Data in A2:A11, the score to rank is in B2
=PERCENTRANK.INC(A2:A11, B2) * 100
' Use PERCENTRANK.EXC to exclude endpoints (0 and 1)
=PERCENTRANK.EXC(A2:A11, B2) * 100
PERCENTRANK.INC includes the minimum and maximum values in the dataset as endpoints. PERCENTRANK.EXC excludes them — no value can return exactly 0 or exactly 1. For most classroom or reporting work, PERCENTRANK.INC is the right choice.
The legacy function PERCENTRANK (without a suffix, available back to Excel 2003) behaves identically to PERCENTRANK.INC.
PERCENTILE.INC and PERCENTILE.EXC return a data value at a given percentile — the inverse operation. If you have a score and want its rank, use PERCENTRANK. If you have a rank and want the corresponding score, use PERCENTILE. Using the wrong function gives a result in entirely different units. Several statistics reference sites make this error; it's the most common Excel mistake in this topic area.
Real-World Applications of Percentile Rank
Percentile rank shows up in more places than exam scores. Here are four common contexts where you'll encounter it.
Standardized testing (SAT, ACT, GRE). Score reports show both the scaled score and the percentile rank. The College Board recalculates percentile ranks annually as the reference population changes, using the extended formula across the full testing population for a given year.
Pediatric growth charts. CDC and WHO charts express children's height and weight as percentile ranks within an age-and-sex reference population. A child at the 85th percentile rank for weight is heavier than 85% of children of the same age and sex in the reference group.
Academic class rank. Some institutions report GPA standing as a percentile rank within a graduating cohort, particularly for medical and law school applications where relative standing matters as much as the GPA itself.
Sports analytics and performance metrics. Athletes' measurable stats — sprint speed, vertical jump, strength ratios — are often reported as percentile ranks within a league or competitive cohort, making cross-player comparisons easier than raw numbers alone.
Need to find a percentile value (the inverse operation) rather than a rank?
Percentile Calculator →The IQR calculator and the five-number summary calculator are natural next steps — Q1 and Q3 are simply the 25th and 75th percentile ranks. To see how those quartiles frame the data visually, the box plot guide walks through reading box-and-whisker plots from scratch.
Common Mistakes
B counts values strictly below the score, not values at or below it. Including the score itself in B inflates the result. Some calculators use the "less than or equal to" convention — always check which one a tool uses before comparing results.
When F > 1, the basic formula PR = (B/N) × 100 ignores the tied values entirely and gives an answer that's too low. If any other score in the dataset equals the score you're ranking, switch to the extended formula: PR = [(B + 0.5 × F) / N] × 100.
The highest score in a closed dataset cannot be above itself. The maximum PR is (N−1)/N × 100. For a dataset of 20 values, the top score has a maximum PR of 95, not 100. Any tool or textbook claiming a top score has PR = 100 is wrong under the strictly-less-than convention.
Percentile rank is a percentage position (output). Percentile is a data value (input). Using PERCENTILE.INC in Excel when you need PERCENTRANK.INC is a common version of this error — the functions return results in completely different units.
Frequently Asked Questions
The basic formula is PR = (B / N) × 100, where B is the number of values strictly below the score and N is the total number of values in the dataset. For datasets with tied values, use the extended formula: PR = [(B + 0.5 × F) / N] × 100, where F is the frequency (count) of the score itself in the dataset.
A percentile is a data value: the 90th percentile is the specific score below which 90% of the data falls. A percentile rank is a percentage position (0–100): a percentile rank of 90 means your score is higher than 90% of the data. A percentile tells you a score; a percentile rank tells you a position.
Step 1: Sort all values in ascending order. Step 2: Count the number of values strictly below your score (B). Step 3: Count the total number of values (N). Step 4: Divide B by N and multiply by 100: PR = (B / N) × 100. For tied values, use: PR = [(B + 0.5 × F) / N] × 100.
Use PR = [(B + 0.5 × F) / N] × 100, where B = number of values strictly below the score, F = frequency of the score in the dataset, and N = total values. The 0.5 × F term gives each tied value equal partial credit. This is the method used in standardized testing (SAT, GRE, ACT score reports).
A percentile rank of 75 means the score is higher than 75% of the other values in the dataset. If you scored in the 75th percentile rank on a test, you performed better than approximately 75 out of every 100 test-takers.
Use =PERCENTRANK.INC(array, x). This returns a value between 0 and 1; multiply by 100 to get a percentage. Use PERCENTRANK.EXC if you want to exclude the endpoints. Important: PERCENTILE.INC and PERCENTILE.EXC return a data value at a given percentile (the inverse operation) — they do not return a percentile rank.
Key Takeaways
The basic formula is PR = (B / N) × 100. B = values strictly below the score, N = total values.
When tied values exist, use the extended formula: PR = [(B + 0.5 × F) / N] × 100. The 0.5 × F term gives each tied value equal partial credit.
Percentile rank ≠ percentile. Percentile rank is a percentage position (output). Percentile is a data value (input). They are inverses of each other.
The maximum PR in a closed dataset is (N−1)/N × 100, not 100. The top score cannot rank above itself.
In Excel, the correct functions are PERCENTRANK.INC and PERCENTRANK.EXC — not PERCENTILE.INC or PERCENTILE.EXC.
B counts values strictly below the score, not at or below it. The "strictly less than" convention is standard in statistics education and standardized testing.
Once you have both formulas — the basic one for clean data, the extended one for ties — the calculation is just counting and dividing. The tricky part is staying clear on what the result means: a position, not a score. Bookmark the statistics formulas page and the statistics cheat sheet for exam prep.
Percentile Calculator · Z-Score Guide · Descriptive Statistics · Percentiles · IQR Calculator · Five-Number Summary Calculator · How to Read a Box Plot · Statistics Formulas · Statistics Cheat Sheet · Study Tips Blog