Excel Formula Checker
Audit the formulas in a workbook without opening Excel. Drop an XLSX, XLSM, XLS or ODS file and every formula on every sheet is listed with its saved value and findings: error results such as #REF! or #DIV/0!, broken references, unbalanced parentheses, unknown function names, circular self-references, links to other workbooks, volatile functions, functions that need Excel 365, approximate VLOOKUPs and hard-coded numbers. You can also check a single formula by pasting it.
- Runs in your browser
- No sign-up
- Free to use
| Cell | Formula | Value | Findings |
|---|
How to use Excel Formula Checker
- Drop a workbook, or switch to “Check one formula”.
- Read the summary of errors, warnings and notes.
- Review the findings per cell.
- Download the report as CSV to fix issues systematically.
Excel Formula Checker features
Every formula
All sheets scanned, with cell address, formula and saved value.
Error explanations
#REF!, #DIV/0!, #NAME?, #VALUE!, #N/A, #SPILL! and more, explained.
Syntax checks
Unbalanced parentheses and quotes; unknown function names.
Risk checks
Circular self-references, external links, volatile functions, approximate VLOOKUP.
Compatibility
Flags functions that only exist in Excel 2021 and Microsoft 365.
Private
The workbook is read on your device; CSV report for follow-up.
When to use Excel Formula Checker
- Reviewing a financial model before sending it to a client.
- Finding #REF! errors after rows or sheets were deleted.
- Checking whether a workbook will work in an older Excel version.
- Auditing spreadsheets inherited from colleagues.
Excel Formula Checker FAQ
Does the tool recalculate the workbook?
No. It reads the formulas and the values Excel saved with the file. Errors are detected from those saved values and from analysing each formula’s text.
What counts as an unknown function?
A name followed by a parenthesis that is not an Excel worksheet function. It may be a typo, an add-in function or a LAMBDA defined in the workbook.
Why are volatile functions flagged?
NOW, TODAY, RAND, OFFSET and INDIRECT recalculate whenever anything changes. In large workbooks that slows everything down; INDEX is often a non-volatile alternative to OFFSET.
What is wrong with VLOOKUP without FALSE?
Without FALSE (or 0) as the fourth argument, VLOOKUP does an approximate match that assumes sorted data and silently returns wrong values otherwise.
How are circular references detected?
A formula whose references include its own cell is flagged. Longer circular chains through other cells are not traced.
Is my workbook uploaded?
No. It is read in your browser with SheetJS.
Auditing spreadsheet formulas
Spreadsheets run a surprising share of business decisions, and their formulas are rarely reviewed as carefully as program code. Studies of real-world workbooks regularly find errors in a large proportion of them: references broken by deleted rows, lookups matching the wrong value, and constants buried inside formulas that nobody updates. A systematic check finds many of these in seconds.
The checker reads every formula stored in the file together with the value Excel calculated when it was last saved. Error values such as #REF!, #DIV/0!, #N/A and #NAME? are reported with an explanation of their usual cause. Each formula is also analysed as text: parentheses and quotation marks must balance, and every function name must exist in Excel, otherwise the formula shows #NAME?.
Some findings are about risk rather than errors. Links to other workbooks break when files move, volatile functions recalculate constantly and slow large models, whole-column references are expensive inside array formulas, and VLOOKUP without an exact-match argument returns plausible but wrong answers on unsorted data. Formulas that reference their own cell create circular references.
Compatibility matters when workbooks are shared. Dynamic array functions such as FILTER, UNIQUE, XLOOKUP and LET, and newer ones such as TEXTSPLIT or GROUPBY, only exist in Excel 2021 and Microsoft 365. The checker points them out so you know which recipients may see #NAME? errors.
Hard-coded numbers inside formulas, deep nesting and very long formulas are reported as notes. They are not errors, but they make models harder to understand and update; moving constants into labelled cells and splitting logic with LET or helper columns improves maintainability. The CSV report lists every finding for follow-up.