n8n Data table nodeConfigure Data table in n8n.

Removes every row matching the conditions you set and returns what it removed. The removal is permanent, so the node gives you a way to look before you leap.

• Data table: choose the table from the list, by name, or by ID.
• Must Match: allConditions narrows the selection, anyCondition widens it.
• Conditions: each one pairs a Column, a Condition and a Value, commonly {{ $json.email }}.
• Dry Run: simulates the delete and returns the affected rows in their before and after states.

Reads rows back out of a table and hands them to the next node as items. This is what turns stored data into workflow data.

• Data table: the table to read from, by list, name or ID.
• Must Match and Conditions: the filter that decides which rows come back.
• Return All: on, every match comes back; off, Limit Per Input Row caps the result for each incoming item.
• Order By, Order By Column and Order By Direction: sort on a column, DESC for newest first.

Acts as a gate on the items coming in. An item whose match is found in the table comes out unchanged; an item with no match produces nothing at all.

• Data table: the table to check against.
• Must Match: allConditions when several columns have to agree.
• Conditions: the column, the comparison and the value, typically {{ $json.id }} taken from the trigger.

The mirror image of the check above. Only items with no match in the table continue, which makes it the shortest deduplication step in n8n.

• Data table: the table holding what has already been processed.
• Must Match: anyCondition is enough when one identifier alone proves the item is known.
• Conditions: usually a single line comparing a column to {{ $json.orderId }}.

Writes a new row for each incoming item. No matching, no conditions: whatever arrives gets a line of its own.

• Data table: the destination table, which must already exist.
• Columns: map the incoming fields onto the table columns, either by hand or automatically when the names match.
• Optimize Bulk: stops the inserted data being returned, which improves bulk insert performance by up to 5 times.

Rewrites the rows that match your conditions with the values you map. Rows that match nothing stay untouched, and no new row is created.

• Data table: the table to modify.
• Columns: the fields to write, so map only what should change.
• Must Match and Conditions: pair a Column, a Condition and a Value such as {{ $json.status }}.
• Dry Run: returns the affected rows before and after, without writing anything.

Updates the matching rows, and inserts a new one when nothing matches. That is a single node where a lookup, a branch and two separate writes used to sit, which is why it usually replaces three nodes in an older workflow.

• Data table: the table that holds the current state.
• Columns: the values to write in both paths, insert and update alike, so map the whole row rather than a fragment of it.
• Must Match and Conditions: the key that decides which branch applies, often {{ $json.email }}.
• Dry Run: shows the before and after states.

Builds a new table and its columns from inside a workflow, rather than by hand in the interface. Handy when a template has to set up its own storage on first run.

• Name: required, the name of the table to create, plain text or an expression.
• Columns: add one entry per column, each with a Name and a Type among boolean, date, number and string.
• Reuse Existing Tables: returns the existing table with the same name instead of raising an error.

Drops an entire table, its columns and its rows in one go. The action cannot be undone, which is why it sits on its own with a single parameter and no filter to soften it.

• Data table: required, identified from the list, by name, or by ID; prefer the ID when two tables share a similar name.

Returns the tables in the project as items, one per table. Useful when a workflow has to work out what exists before it writes anything.

• Return All: on for the whole list; off, and Limit Per Input Row caps it, for example at 50.
• Filter by Name: keeps the tables whose name contains the text, matching without regard to case.
• Sort Field and Sort Direction: sort on createdAt, name or updatedAt, ascending or desc.

Changes the name of an existing table. Nothing else about the table moves: columns, types and rows all stay exactly as they were.

• Data table: required, the table to rename, from the list, by name or by ID.
• New Name: required, the new name, often built from an expression such as {{ $json.period }}.