How to Import CSV Data in Power BI (Clean Import Guide) Easily 2026

Check the file for consistent formatting first: headers in the first row, no merged or blank header cells, and one data type per column. Mixing text and numbers in the same column is one of the most common causes of a messy import.

If the file came from another system, open it once in a plain text editor or spreadsheet app to confirm the delimiter is a comma, not a semicolon, which is common in regions that use commas as decimal separators and can make Power BI read an entire row as a single unsplit column. Trim any trailing blank rows too; Power BI imports them as rows of nulls that need cleaning up later.

Launch Power BI Desktop and its start screen shows a Blank report tile, a Recommended section, and a list of recently opened files. Choose Blank report for a new project, so the CSV lands in a clean canvas instead of an existing file's data model. Recent .pbix files are listed below with their save location, useful when continuing earlier work.

From the ribbon, click Get data and look under Common data sources for Text/CSV, sitting alongside Excel workbook, SQL Server, and Web. This is the connector Microsoft's own documentation lists for delimiter-separated text files, distinct from Import or Excel workbook, which expect different file structures. Selecting Text/CSV opens a local file browser, not a data-source login screen, since a local file doesn't need credentials.

In the Open dialog, navigate to wherever the file lives, select it, and click Open. Power BI opens the Navigator with a preview of the detected delimiter, column headers, and inferred data types.

From here, Transform Data opens the Power Query editor for cleanup before anything loads; Load brings the data straight into the report. If a column's type looks wrong in the preview, use Transform Data and fix it there. Power Query also allows choosing whether type detection scans the first 200 rows or the full file, or turning it off so every column defaults to Text, useful when a numeric-looking column is really an ID that shouldn't be summed.

Check the imported table against what was expected: column names should match the source headers, and numeric columns should show as numbers, not text, confirmed if a Sum or Total calculates correctly. A quick sanity table with a running total from three or four rows shows in seconds whether the import is clean or needs a trip back through Power Query.

One limitation worth knowing before relying on this file long-term: Power BI Desktop records a fixed column count from the initial import. If the source CSV later gains new columns and the report is refreshed, those new columns won't appear. Editing the query's Source step, or reimporting the file, fixes it; refresh alone won't pick up structural changes.

Once an import works, note which connector, delimiter, and column fixes were used, so the next file from the same source goes faster. Power Query's applied-steps list saves those transformation steps, so pointing the same query at a new file becomes close to a one-click refresh instead of a re-diagnosis every time.

A clean CSV import doesn't have to stand alone either. Power BI can relate it to tables from Excel, SQL Server, or any other connector in the same model, provided there's a shared key column between them. This is where the earlier formatting check pays off: a sales export sitting next to a product catalog only joins cleanly if both sides' types and values line up.