Fix "Expression.Error: The column of the table wasn't found"

A Power Query step names a column that isn't there any more. Read the column name in the error, find the step that uses it, then fix the step or let it skip.

Power BIFix a problem8 min read

Checked against Microsoft Learn 24 Oct 2026 · 10 sourcesPosted by the Maker DeskHow we write guides

A Power BI refresh, or a query in Power Query Editor, stops with "Expression.Error: The column 'Date' of the table wasn't found." The data source still works and nothing is corrupted. One step in your query refers to a column by name, and when it ran, the table it received had no column with that name. Power Query doesn't guess, so the step fails and the query stops loading. The fix is to find that step and bring it in line with the columns the table has today.

Why Power Query says a column of the table wasn't found. A query's applied steps run in order: Source reads the file lowcodestacks-orders.csv, Promoted Headers turns its first row into column names, and Changed Type sets the type of a column named Date. The file's header now says Order date instead of Date, so Changed Type can't find Date, and the query stops with Expression.Error: The column 'Date' of the table wasn't found, with Details: Date. The steps after it, such as Renamed Columns, never run. A name goes missing for one of three reasons: it was renamed or removed at the source; the headers were promoted differently, so the columns are called Column1, Column2 and so on; or an earlier step in the same query renamed or removed it. The fix: point the step at today's column name or delete the step, use MissingField.Ignore where a column is allowed to be absent, and republish from Power BI Desktop. — A step asks for a column by name; when the table no longer has it, the query stops at that step.

View full size

Power Query Editor shows a yellow bar under the formula bar: Expression.Error: The column '…' of the table wasn't found., with the column name under Details.
In Power Query Online and dataflows, the same error also shows ErrorCode = 10224.
A scheduled refresh in the Power BI service fails after someone changed the source file, sheet or table.
It names a column such as Column1 that you never typed yourself.

What the error means

Microsoft splits a step-level error into three parts:

Part	In this error	What it tells you
Error reason	`Expression.Error`	The step's formula failed
Error message	`The column 'Date' of the table wasn't found.`	Which kind of failure: a column name the table doesn't have
Error detail	`Date`	The exact name the step asked for

Microsoft's own explanation is short: a step makes a direct reference to a column name that doesn't exist in the query. A step-level error stops the whole query from loading, so the steps after it never run, and nothing reaches the report.

The Power Query M language is case-sensitive, so a step that asks for date won't find a column called Date. A column can look present in the preview and still fail the lookup.

Find the step and the name

In Power BI Desktop, select Transform data to open Power Query Editor, and select the query that fails.
In the list of applied steps, select each step from the top. The first one that shows the yellow error bar is the step to fix. If the bar offers Show details, select it to see the reason, the code and the detail.
Read the step's formula in the formula bar. The missing name is in it, for example Table.TransformColumnTypes(#"Promoted Headers", {{"Date", type date}}).
Select the step just before it and look at the column headers in the preview. That's the table the failing step received. Compare its names with the one in the formula.

Microsoft's example of this error shows the pattern in reverse: a step renames a column called Column to Date, but the source file was changed by hand and already has a Date heading, so there's no Column left to rename.

Why the column is missing

What happened	How to spot it	Fix
The column was renamed or removed at the source: a new export, an edited header, a changed view	The step before the failing one shows the new name, or no such column	Change the name in the step's formula to today's name, or remove the step if the column is gone for good
The headers were promoted differently, so columns are named `Column1`, `Column2` and so on	The error names `Column1`, or a real header such as `Customer` when the preview shows `Column1`	Fix the promote-headers step: a change at the source (an extra title row, say) can move the real headers down a row
An earlier step renamed or removed the column, often after steps were reordered or one was edited	The column exists after Source but disappears a few steps later	Point the failing step at the new name, or move it before the rename
An automatic Changed Type step lists every column by name	The failing step is Changed Type right after Promoted Headers	Edit or delete the step, then set types again for the columns you have

The last one is common with files. For unstructured sources such as Excel, CSV and text files, Power Query inspects the first 200 rows and adds two steps automatically: one that promotes the first row to column headers, and Changed type, which names each column it typed. So when a later file arrives with different headers, that step is often the first to fail. The behaviour is set in Power BI Desktop under File > Options and settings > Options, in the Type Detection section of Data Load:

under CURRENT FILE, the check box Detect column types and headers for unstructured sources;
under GLOBAL, a choice between always, never, and according to each file's setting.

Let a step skip a missing column

Some columns come and go legitimately, such as an optional column in a monthly file. Several table functions take an optional last argument that says what to do when a column is missing, instead of raising the error:

Value	What happens
`MissingField.Error`	The default: the step fails with this error
`MissingField.Ignore`	The missing column is skipped
`MissingField.UseNull`	The missing column is added, filled with null

Table.RenameColumns, Table.RemoveColumns and Table.SelectColumns take it as the third argument:

powerquery

Table.RenameColumns(
    #"Promoted Headers",
    {{"Order date", "Date"}},
    MissingField.Ignore
)

Table.TransformColumnTypes takes it inside a record in its third argument, beside the culture:

powerquery

Table.TransformColumnTypes(
    #"Renamed Columns",
    {{"Date", type date}, {"Discount", type number}},
    [Culture = "en-GB", MissingField = MissingField.UseNull]
)

Warning

Use this only for columns that are truly optional. If the column your visuals depend on disappears, skipping it hides the problem: the refresh succeeds, and measures that use the column fail or show blanks instead.

Publish the fix

Power BI Desktop and the Power BI service treat a changed column differently. Microsoft's refresh documentation says a data refresh in the service fails when a source column or table is renamed or removed, because the service doesn't do a schema refresh. So:

Fix the query in Power BI Desktop, apply the changes so the model reloads, and check the visuals that use the column.
Publish the semantic model again, then refresh it in the service.

A schema refresh also removes the old column from the model, so measures, relationships and row-level security that used it may need fixing too.

Read the name under Details first; it tells you exactly what to search for.
Rename columns once, in one step near the start, and refer to the new names after it.
Agree with whoever produces the file that header changes are announced.

Don't

Don't delete every step after the failing one to make the error go away.
Don't use MissingField.Ignore on a column the report needs.

If the error names a table rather than a column, see the key didn't match any rows in the table. For other ways a refresh can fail, work through our Power BI refresh checklist.