R-squared says how much of the variation in y is explained by x — from 0 (no relationship) to 1 (a perfect line). It’s the headline number for any trend or regression.
RSQ = CORREL^2. 0.85 means x explains 85% of y’s variation.
The example
The fit quality of a trend.
| A | B | |
|---|---|---|
| 1 | Measure | Value |
| 2 | Correlation | 0.92 |
| 3 | R-squared | 0.85 |
The formula
The formula:
How it works
How it works:
RSQ(known_ys, known_xs)returns a value from 0 to 1.- It equals the square of the correlation:
RSQ = CORREL(ys,xs)^2. - Read it as a percentage: 0.85 means the line explains 85% of the variation in y.
- Higher is a tighter fit — but a high R-squared doesn’t prove causation, only association.
High R-squared ≠ good model. It can be inflated by outliers, a curved relationship forced into a line, or coincidental correlation. Always look at a scatter plot alongside the number.
Try it: interactive demo
Pairs “x,y”.
Variations
From correlation
Square it:
Percent form
As a %:
Adjusted R²
From LINEST stats.
Pitfalls & errors
Not causation. R-squared measures fit, not cause — correlated isn’t the same as causal.
Outliers inflate it. A couple of extreme points can prop up R-squared; check the scatter.
Linear only. RSQ measures linear fit; a strong curve can show low R-squared.
Practice workbook
Frequently asked questions
How do I calculate R-squared in Excel?
What is a good R-squared value?
Does a high R-squared prove x causes y?
Stop fighting formulas. Learn them in a day.
This recipe is one of hundreds of real-world formulas we teach. Our Excel Formulas & Functions class covers lookups, logic, text, and dynamic arrays hands-on — live in Dallas–Fort Worth, Houston, Austin, Oklahoma City, Denver, or online.
See the Formulas & Functions Class