Trimmed Mean Calculator
Calculate the trimmed mean of any dataset by removing a chosen percentage of the smallest and largest values — a robust average that shrugs off outliers. Paste your numbers, pick a trim level, and see the full step-by-step solution alongside the ordinary mean and median.
Textbook-accurate method · Reviewed July 2026 · Uses the standard per-side truncated-mean definition
Enter Your Data
Works with commas, spaces, tabs, or line breaks. Decimals and negatives are fine. Values are sorted automatically.
This is the percentage removed from each end (the standard statistical definition). A 10% trim removes the lowest 10% and highest 10%. Note: Excel’s TRIMMEAN uses the combined total instead.
Your Result
Trimmed Mean vs. Other Averages
What Gets Trimmed
Grey points are removed; teal points are kept
- Kept
- Trimmed
- Trimmed mean
What Is a Trimmed Mean?
A trimmed mean (also called a truncated mean) is an average calculated after removing a set percentage of the smallest and largest values from a dataset. By cutting away the extremes before averaging, it protects the result from being dragged around by outliers. The trimmed mean sits neatly between two familiar measures: the ordinary mean, which uses every value and is fully exposed to outliers, and the median, which effectively trims everything except the middle. A trimmed mean keeps most of your data while discarding just the unreliable tails.
You’ve almost certainly seen a trimmed mean in action without realising it. In Olympic events like gymnastics and diving, the highest and lowest judges’ scores are thrown out before averaging the rest — that’s a trimmed mean, and it stops a single biased or mistaken judge from swinging the outcome.
How to Calculate a Trimmed Mean
The method is straightforward, and the calculator above runs every step for you:
- Sort the data from smallest to largest.
- Decide the trim percentage for each end — commonly 5%, 10%, or 20%.
- Work out how many values to cut from each end: multiply the count by the trim proportion and round. For a 10% trim on 10 values, that’s 10 × 0.10 = 1 value from each side.
- Remove those values from both the top and bottom.
- Average what remains — that’s your trimmed mean.
The trimmed mean formula
If k is the number of values trimmed from each end and the sorted data is x₍₁₎, x₍₂₎, …, x₍ₙ₎, the trimmed mean is:
Trimmed mean = ( x₍ₖ₊₁₎ + x₍ₖ₊₂₎ + … + x₍ₙ₋ₖ₎ ) / ( n − 2k )
where k = round(n × p) and p is the trim proportion for each end. In words: add up all the values left after trimming, then divide by how many remain.
A worked example
Take the dataset 12, 15, 18, 20, 22, 24, 25, 26, 28, 55 — notice that 55 is a clear outlier. The ordinary mean is 24.5, pulled upward by that one extreme value. For a 10% trimmed mean, we cut one value from each end (the 12 and the 55), leaving 15, 18, 20, 22, 24, 25, 26, 28. Their average is 22.25 — a truer picture of the typical value, no longer distorted by the outlier. That single number is exactly what the calculator above produces, along with a visual of which points were removed.
Trimmed Mean vs. Mean vs. Median
These three measures of central tendency form a spectrum of robustness — how much they resist being distorted by extreme values:
| Measure | How it handles outliers | Best when |
|---|---|---|
| Mean | Fully exposed — every value counts | Data is clean and symmetric |
| Trimmed mean | Robust — extremes removed, rest kept | A few outliers, but tails still informative |
| Median | Most robust — only the middle matters | Heavily skewed data or many outliers |
The trimmed mean is often the sweet spot: more stable than the mean, but using more of the data than the median. For a full picture of a dataset including the mean, median, and quartiles, try our Descriptive Statistics Calculator.
Trimmed Mean vs. Winsorized Mean
A close cousin of the trimmed mean is the Winsorized mean. Instead of deleting the extreme values, it replaces them with the nearest remaining value, then averages the whole set. Trimming throws the tails away; Winsorizing pulls them in. The Winsorized mean keeps the same number of data points as the original, which can be preferable in some statistical procedures. The calculator above reports the Winsorized mean alongside the trimmed mean so you can compare the two approaches at a glance.
A note on the trim percentage: standard vs. Excel
There’s a common source of confusion worth flagging. The standard statistical definition — used by this calculator and most textbooks — treats the trim percentage as the amount removed from each end. So a 10% trim removes 10% from the bottom and 10% from the top, 20% in total. Excel’s TRIMMEAN function, by contrast, treats the percentage as the combined total across both ends, so you’d enter 20% in Excel to match a 10% per-side trim here. If your numbers don’t line up with a spreadsheet, this is almost always why.
Where the Trimmed Mean Is Used
- Sports judging: Olympic gymnastics, diving, and figure skating drop the highest and lowest scores.
- Finance: smoothing volatile stock returns so a single extreme day doesn’t distort the average.
- Economics: central banks publish “trimmed mean” inflation measures that strip out the most volatile price movements.
- Science & engineering: reducing the impact of measurement errors and faulty sensor readings.
- Survey research: limiting the influence of a few extreme responses on the overall average.
Frequently Asked Questions
A trimmed mean is an average calculated after removing a fixed percentage of the smallest and largest values from a dataset. Removing the extremes makes it far less sensitive to outliers than the ordinary mean, while still using more of the data than the median. It’s also called a truncated mean, and it’s the method behind Olympic scoring, where the highest and lowest judges’ scores are dropped.
Sort the data from smallest to largest, decide a trim percentage for each end, and multiply the number of values by that proportion to find how many to remove from each side (rounding to a whole number). Remove that many values from both the top and bottom, then take the ordinary average of what remains. For example, a 10% trim on 10 values removes one from each end and averages the middle eight.
Under the standard statistical definition, a 10% trimmed mean removes the lowest 10% and the highest 10% of the values — 20% of the data in total — then averages the remaining 80%. This calculator uses that per-side convention. Be aware that Excel’s TRIMMEAN function interprets the percentage as the combined total across both ends, so you would enter 20% in Excel to get the same result.
Use a trimmed mean when your data contains outliers or a few extreme values that would distort the ordinary average, but you still want to use most of the data rather than reducing it to a single middle value like the median. It’s ideal for skewed data, noisy measurements, volatile financial returns, and any situation where a small number of extremes shouldn’t dominate the result.
Both reduce the influence of outliers, but they handle the extremes differently. A trimmed mean deletes the extreme values entirely and averages what’s left. A Winsorized mean replaces the extreme values with the nearest remaining value instead of deleting them, then averages the full set, so it keeps the same number of data points. Trimming discards the tails; Winsorizing caps them.
Common choices are 5%, 10%, and 20% per side, with 10% being a popular default. The right amount depends on how many outliers you expect: light trimming (5%) handles occasional extremes, while heavier trimming (20–25%) suits noisier data. Trimming 25% from each end produces the interquartile mean, which averages only the middle half. Avoid trimming so much that too little data remains to be meaningful.
Methodology & formulas used
This calculator uses the standard per-side truncated-mean definition, reviewed July 2026. All computation runs locally in your browser; your data is never uploaded.
- Values trimmed per end: k = round(n × p), where p is the per-side trim proportion
- Trimmed mean: the arithmetic mean of the values remaining after removing k from each end
- Winsorized mean: the k extreme values on each end are replaced by the nearest kept value, then the full set is averaged
- Ordinary mean: Σxᵢ / n, shown for comparison
- Median: the middle value (or average of the two middle values), shown for comparison
The trim percentage here is applied to each end, matching the standard statistical convention used by most textbooks and calculators. Excel’s TRIMMEAN function instead treats the percentage as the combined total across both ends and rounds the trimmed count down to an even number, so results can differ. Trimming must leave at least one value; if 2k is greater than or equal to n, the trimmed mean is undefined. Results are rounded for display but computed at full precision.
References
- Wikipedia. Truncated Mean (Trimmed Mean) — definition and properties. en.wikipedia.org
- Statistics By Jim. Trimmed Mean: Definition, Calculating & Benefits. statisticsbyjim.com
- Statology. How to Calculate a Trimmed Mean (Step-by-Step). statology.org