Create a Dataverse Table from Excel or CSV in Power Apps (Step by Step)

The SPGuides purchasing team tracked eight vendors in an Excel workbook that got emailed around for every update. They wanted it in Dataverse, with no retyping.

In this tutorial, we’ll create Dataverse table from Excel data in Power Apps: upload the workbook, check the AI column types, fix the table properties, then import more rows from a second file.

Here’s what we’re building: a Vendor table with all 10 vendors, eight from the first file and two added later.

Before You Start

A Dataverse table is a database-backed sheet with real data types and security. New to it? Start with what Microsoft Dataverse is.

You need System Customizer to create tables and Environment Maker to import rows. Get Data from Excel isn’t in the Power Apps Developer Plan.

Our workbook, SPGuides Vendors.xlsx, has a Vendors sheet with seven columns:

Excel column	Example value	Dataverse type the AI picked
Vendor Name	Fabrikam Inc	Text (became the primary column)
Contact Name	Megan Bowen	Text
Email	A work email address	Email
Phone	A business phone number	Phone
State	WA	Text
Contract End Date	12/31/2026	Date only
Active	Yes	Yes/No

The eight sample rows (Email and Phone left out):

Vendor Name	Contact Name	State	Contract End Date	Active
Fabrikam Inc	Megan Bowen	WA	12/31/2026	Yes
Northwind Traders	Alex Wilber	WA	03/31/2027	Yes
Adatum Corporation	Adele Vance	IL	11/30/2026	Yes
Litware Inc	Lee Gu	TX	06/30/2027	Yes
Tailspin Toys	Diego Siciliani	MA	10/31/2026	No
Wingtip Toys	Grady Archie	CO	01/31/2027	Yes
Proseware Inc	Isaiah Langer	GA	09/30/2027	Yes
Relecloud	Lynne Robbins	NC	12/15/2026	No

Keep one header row and no merged cells or blank rows. Building by hand instead? See the other ways to create a Dataverse table.

Step 1: Upload the Excel File

1. In make.powerapps.com, pick your environment and select Tables.
2. Select the Create with Excel or .CSV file tile.

1. In Import an Excel or .CSV file, select Select from device.
2. When loading finishes (about 10 seconds), check that the Vendors sheet has Include on, then select Import.

Checkpoint: the data workspace shows a table card named Vendor tagged Imported with AI.

Step 2: Review the Column Types the AI Picked

Copilot guesses names and data types from your data, and Microsoft says the guesses might not be 100% accurate.

1. Select View data and keep First row as headers on. Avoid Remove: it deletes a new table and its data.
2. Check each header icon. To change a type, select the header and edit the column.

The preview shows only the first 20 rows; the rest load on save. Active became a Dataverse Yes/No column.

Checkpoint: every column has the type you want, and all eight rows show.

Step 3: Fix the Table Name, Primary Column and Schema

1. Select Properties to open Edit table.
2. The AI set Plural name to Vendor, the same as the display name. Change it to Vendors.
3. Set Primary column to Vendor Name.
4. Under Advanced options, check the Schema name (ours: cr85a_vendor) and select Save.

The schema name can’t change after saving. For your own prefix instead of cr85a, use a solution, as in our Dataverse solution guide.

Checkpoint: the table card still says Vendor, with plural Vendors.

Step 4: Save and Exit to Create the Table

Nothing exists in Dataverse until now. Select Save and exit, then Save and exit again in the Done working? dialog.

After 30 to 60 seconds, the table showed under Tables > Custom.

The AI also chose the logical names you’ll use in flows and formulas:

cr85a_vendorname, cr85a_contactname, cr85a_emailaddress, cr85a_phonenumber,
cr85a_state, cr85a_contractenddate, cr85a_isactive

Checkpoint: Vendor is a Standard table with primary column Vendor Name and 8 rows.

Step 5: Import More Rows from a Second Excel File

Two new vendors arrived in SPGuides New Vendors.xlsx. On the Vendor table, select Import > Import data from Excel.

1. Upload the file, pick the New Vendors sheet and select Next.
2. Leave Primary key as (None); the file holds only new rows.
3. On Possible match, Contract End Date and Active showed “Mismatched column metadata, but matching row-level data.” Select Accept match on both.
4. With Mapped at 7, select Import.

The 19 unmapped columns are system columns. More ways to add rows: adding data to a Dataverse table.

Checkpoint: the dialog closes and you’re back on the table page.

Test the Import

Rows load in the background. An instant refresh still showed 8; two minutes later it showed 10, including Coho Winery and Trey Research, with no duplicates.

To use the rows in a canvas app, see getting data from Dataverse in Power Apps.

Need a Schedule or a Big File? Use a Dataflow

For a file that keeps changing or needs cleanup, pick Import > Import data with Dataflows: a Power Query job that can refresh on a schedule.

Choose Excel workbook or Text/CSV, shape the data, and map it.

Our Dataverse dataflow walkthrough covers every screen. For row-by-row logic, try importing a CSV into Dataverse with Power Automate.

Which Method Should You Use?

Goal	Method	Why
New table from a workbook	Create with Excel or .CSV file	AI builds the columns and loads the rows
Add rows to an existing table	Import data from Excel	Column mapping, optional key to update rows
Repeat the load on a schedule	Import data with Dataflows	Power Query shaping and refresh
Custom logic per row	Power Automate flow	Lookups, conditions, notifications

Troubleshooting and Things to Keep in Mind

A date looks a day off in a flow. The AI made Contract End Date a Date only format with User local behavior. The grid showed 11/30/2026, but the raw Web API value fell on the previous day in UTC. Switch the behavior to Date only right away; the one-time change affects only values saved afterward.

The Import button is grayed out. A Possible match is still open. Accept or reject each one.

“You do not have permission to access these records.” I got this opening the table via a main.aspx link with no app. Add the table to an app.

Repeat imports duplicate rows. With Primary key (None), every row is added as new. Pick a key, such as an alternate key, to update matching rows.

Some columns won’t import. Timezone, multiselect Choices, Image and File columns aren’t supported. The uploaded workbook stays in the File Upload table until you delete it.

Where to Take It Next

• Put the table in front of users with a model-driven app in Power Apps.
• Give each vendor a readable ID with an autonumber primary column.
• Share the data back with an export of the Dataverse table to Excel.

More Dataverse builds are on our Power Apps tutorials hub.

If you’re moving more data into Dataverse, these are good next reads:

• SharePoint List to Dataverse: Create a Dataverse Table From a SharePoint List (4 Ways)
• How to Create Dataverse View [With Examples]
• How to Delete All Records From a Dataverse Table (4 Tested Ways)
• Power Automate: Add a New Row to Dataverse (Lookups, Choices and the New Row ID)

Frequently Asked Questions