In Excel: use =LINEST(known_ys, known_xs, TRUE, TRUE) — it returns the regression coefficients as an array, plus R², standard errors and the other statistics when stats is TRUE.
On this page8
Syntax
Arguments
| Argument | required / optional | Description |
|---|---|---|
known_ys | required | The outcomes being modelled. |
known_xs | optional | One or more columns of predictors. |
const | optional | FALSE forces the intercept to zero. |
stats | optional | TRUE returns the full diagnostic block including R² and standard errors. |
Related functions
Select a cell with several empty rows and columns beneath and to the right.
Type =LINEST( and select the y column, then the block of x columns.
Add , TRUE, TRUE to fit an intercept and return the diagnostics.
Press Enter in Excel 365, or select a 5-row block and press Ctrl+Shift+Enter in older versions.
What this does
LINEST is Excel's full least-squares regression in a single function. With the fourth argument set to TRUE it returns a block of diagnostics rather than just coefficients: R², standard errors for each coefficient, the F statistic and the degrees of freedom. Two things surprise everyone the first time. The coefficients come back in reverse order, with the last predictor leftmost and the intercept at the far right. And it accepts several predictor columns, which makes it a genuine multiple regression without the Analysis ToolPak. In Excel 365 the block spills; older versions need a selected range and Ctrl+Shift+Enter. Most people learn this as a sequence of clicks and forget it by next week; learning it as a pattern instead is what lets you apply it to the next, slightly different version of the problem without starting from scratch. That is the difference this page is trying to make. The difference between a quick fix and a sheet you can trust is the extra minute you spend validating “perform multiple regression in excel”. Start on a copy or a tiny sample, keep the affected cells visible, and compare the result with the tool above before you touch the real workbook. When a formula is involved, keep the inputs labelled beside it, reference cells instead of typing values, and apply number formatting only after the result checks out. The point is a text operation that turns messy entries into clean, usable data, but the practical win is that someone else can open the file and understand what happened without asking you.
A worked example
Sales in D2:D100 modelled on spend in B2:B100 and headcount in C2:C100: =LINEST(D2:D100, B2:C100, TRUE, TRUE) spills a five-row block. The top row holds the coefficients right-to-left (headcount, spend, intercept) and the third row's first cell is R². For a single predictor, SLOPE and INTERCEPT are far easier to read. LINEST is a complete regression without a add-in, which matters whenever the model has to live in the workbook rather than in a separate tool. If there is any chance you will reuse this, drop it into a small template tab right now: a labelled input area on the left and the formula beside it, checked once against the tool above. Next time the same question comes up, the answer is a single paste away instead of a rebuild from memory.
In Google Sheets
Google Sheets handles this almost identically to Excel. The formula syntax above is the same, and the menu lives under a slightly different label rather than a ribbon tab. Use the platform toggle at the top of the page to switch every keyboard shortcut between Windows and Mac, and expect at most cosmetic differences in naming. Keep this page bookmarked for the next time the same question comes up. Better still, repeat the steps once in your own workbook — doing it yourself is what turns a copied answer into something you remember. If you take one thing from this page on “perform multiple regression in excel”, make it the habit rather than the keystrokes: set the problem up with labelled inputs, reference those cells, and let Excel do the recomputing. Bookmark the page for the syntax, but do the example once in a blank sheet and check it against the tool above — five minutes of hands-on practice fixes the method in memory far better than re-reading, and it surfaces the small snags while they are still harmless. After that the technique is genuinely yours: faster than searching for it again, and reliable enough to drop into work that other people depend on.
Common mistakes
- Reading the coefficients left to right — they are returned in reverse predictor order, with the intercept last.
- Omitting the final TRUE and then looking for R², which is only returned when stats is switched on.
- Adding predictors without checking they are not collinear, which makes individual coefficients unstable and misleading.
Frequently asked questions
How do I read the LINEST output?
Row 1 holds the coefficients in reverse order of the x columns, with the intercept rightmost. Row 2 holds their standard errors, and row 3 column 1 is R².
Can LINEST do multiple regression?
Yes — pass several adjacent columns as known_xs and it fits them all.
Where is R² in the output?
The first cell of the third row, and only when the fourth argument is TRUE.