Percent Difference Calculator

How to Use This Calculator

1. Enter Value 1 and Value 2 — order does not matter.
2. Read the symmetric percent difference instantly.
3. Use vs Percent Change tab to compare both formulas side-by-side and understand the difference.
4. Use Multiple Pairs tab to compute percent difference for 3 pairs at once.
5. Professional tier adds log difference and fractional difference alternatives.

Formula

Example

Frequently Asked Questions

• What's the difference between percent difference and percent change? ▼
  Percent difference and percent change are similar in name but answer different questions. Percent change is directional — it asks 'by what percent did value A change when it became value B?' The formula is (B − A) / |A| × 100%, where A is the reference starting point. Swapping A and B gives a different answer: going from 80 to 100 is +25% change, but going from 100 to 80 is −20% change. Percent difference, by contrast, treats both values symmetrically and asks 'how different are these two values relative to their average?' The formula is |A − B| / ((A + B) / 2) × 100%. Comparing 80 and 100: percent difference = |80 − 100| / 90 × 100% = 22.2%, regardless of which is first. This symmetry makes percent difference appropriate when there is no natural starting point or direction — comparing the size of two cities, two lab measurements, two groups' scores — where you would not call one value before and the other after. Percent change is appropriate for time-series data, growth rates, and before/after scenarios where direction matters.
• When should I use percent difference instead of percent change? ▼
  Use percent difference when neither value has special status as a reference or starting point. The clearest signal is when swapping the two values should give the same answer — if the direction of comparison is arbitrary, use percent difference. Scientific examples: comparing the density measured by two different methods (neither is more before than the other), comparing the performance of two sports teams in the same season, comparing two students' test scores, or comparing the price of the same item at two different stores. Business examples: comparing revenue of two divisions in the same quarter, comparing the output of two machines on the same day, comparing two candidates' polling numbers. If you have a baseline, standard, or starting point, use percent change instead. If one value is a known accepted standard and the other is a measurement, use percent error. A common mistake is using percent change when the values have no temporal relationship, which gives a misleading directional framing. Saying 'City A's population is 25% higher than City B's' implies A is the reference; saying the populations differ by 22.2% is symmetric and neutral.
• Why is percent difference always symmetric (a vs b = b vs a)? ▼
  Percent difference is symmetric because its formula uses the average of the two values as the denominator: |A − B| / ((A + B) / 2) × 100%. Swapping A and B changes neither the numerator (|B − A| = |A − B|) nor the denominator ((B + A)/2 = (A + B)/2). So the result is identical regardless of order. This symmetry is the whole point — it reflects the absence of a preferred reference value. Compare this to percent change, (B − A) / |A| × 100%, where the denominator is specifically A (the starting value). Swapping gives (A − B) / |B| × 100%, which uses B as denominator, producing a different result. The asymmetry of percent change is also intentional — it correctly represents the directional nature of change from A to B. A memorable example: 80 vs 100 gives 22.2% difference regardless of order. But 80 to 100 is +25% change (using 80 as base), while 100 to 80 is −20% change (using 100 as base). These different denominators explain all three different numbers from the same two values.
• How is percent difference calculated when both values are negative? ▼
  Percent difference uses the absolute value of the average as its denominator, so it handles negative values correctly as long as the average is not zero. For two negative values like −80 and −100: |A − B| = |−80 − (−100)| = |20| = 20. Average = (−80 + (−100)) / 2 = −90. |Average| = 90. Percent difference = 20 / 90 × 100% = 22.2%. The same result as with +80 and +100 — correctly so, since the relative difference is the same. A problematic case: when one value is positive and one is negative with the same magnitude, say +50 and −50, the average is zero, making the denominator zero and percent difference undefined. In this case you cannot use percent difference meaningfully — consider reporting the absolute difference (100) or using a different normalization. When values have opposite signs but non-zero average (like −30 and −70, average = −50), percent difference = |−30 − (−70)| / |−50| × 100% = 40/50 × 100% = 80%, which is well-defined.
• What is the maximum possible percent difference between two values? ▼
  Percent difference has no theoretical maximum — it can exceed 100% and approach infinity in certain cases. When one value approaches zero while the other is nonzero, the average approaches half the nonzero value, but the numerator approaches the nonzero value, so percent difference approaches 200%. For example: A = 0.001, B = 100. Average = 50.0005. |A − B| = 99.999. Percent difference is approximately 99.999 / 50.0005 × 100% = 199.998% — very close to 200% but never quite reaching it for nonzero A. The theoretical maximum approaches 200% asymptotically as one value approaches zero. This is a key difference from percent error (which can exceed 200% easily — an experimental value of 5 vs theoretical of 1 gives 400% error) and from percent change (which can be any positive percentage for increases, and ranges from −100% to 0% for decreases). For values with opposite signs, percent difference can exceed 200%. Example: A = 100, B = −90. Average = 5, |difference| = 190. Percent difference = 190/5 × 100% = 3800%.

Related Calculators