BigQuery Column Aliases and Cross-Region Table Copy Now GA

BigQuery column aliases let you reuse an alias created in the SELECT clause later in the same SELECT list and in the WHERE clause. Google Cloud announced the general availability (GA) of this feature and of cross-region table copy in its October 15, 2026 release notes (BigQuery release notes). If you query GA4 export data in BigQuery, you can drop the habit of wrapping the same formula in a subquery. If you need to move a dataset to another region, recheck the copy method and its limits.

What changed with BigQuery column alias reuse

In a spreadsheet, one cell can use the result of the cell before it. Queries can now do the same: you name a calculation early and call it by name later. Until now, you could not use an alias you had just created in the same SELECT list. You had to copy and paste the formula or wrap the query in one more subquery. Queries migrated from other data warehouses often use this syntax as is, and Google cited migration support as a reason.

The example in the official docs is below. It calls gross_price again in the next item and in the WHERE clause (reusable column aliases docs).

SELECT
  net_price * 1.23 AS gross_price,
  gross_price * 0.5 AS discounted_gross_price
FROM product
WHERE discounted_gross_price < 10.0;

Aliases are referenced only from left to right. If you call an alias before its definition, you get a Column not found error. If you define the same alias twice in one SELECT list, you get a Column name is ambiguous error.

How to turn on column aliases: off by default

This feature is off by default. An admin must set the reusable_column_aliases_mode option to on before you can use it. The required permission is roles/bigquery.admin, or bigquery.projects.update at the project level and bigquery.organizations.update at the organization level (how to enable).

Scope	Command
Project	ALTER PROJECT PROJECT_ID SET OPTIONS (`region-REGION.reusable_column_aliases_mode` = "on");
Organization	ALTER ORGANIZATION SET OPTIONS (`region-REGION.reusable_column_aliases_mode` = "on");
Turn off	Change the value to off. NULL means the system default (off).

Put the region where your data lives in REGION, and your project ID in PROJECT_ID. After you turn it on, query the PROJECT_OPTIONS, ORGANIZATION_OPTIONS and EFFECTIVE_PROJECT_OPTIONS views to confirm the value is applied as on.

Can my query results change: check for name conflicts

When you turn aliases on, the results of queries with overlapping names can change. BigQuery looks up a name in this order: table columns, aliases defined in the query, columns of the outer query, function arguments, and script variables. So if you already use an alias with the same name as a DECLAREd variable, the alias value is used instead of the variable once the feature is on. The official docs also tell you to assess existing workloads before you enable it.

There are limits on usage too. An alias that points to a grouping column, or an alias that references an aggregate or window function, cannot be used in the WHERE clause or as an argument of another aggregate function. Pipe syntax and GQL do not support it.

Check with a small experiment. Turn it on in a development project first. Run a few queries whose names overlap with variables or outer query columns before and after, and compare the results. If they match, expand to the production project.

BigQuery cross-region table copy GA: method and limits

Cross-region table copy moves tables, table snapshots and table clones from one region (or multi-region) to another. A typical case is merging a dataset in Seoul (asia-northeast3) into the same place as the US multi-region. The method is the bq cp command, and the basic form is below (table management docs).

bq cp --sync=false mydataset_us.mytable mydataset_eu.mytable2

You cannot copy across regions in the console or with the CREATE TABLE ... COPY statement. The other limits in the same doc are below.

Item	Behavior
Policy tags and table tags	You cannot copy if the source has them.
IAM policy	Not copied, so you must apply it again to the destination table.
Time travel data	Not copied to the destination region.
Copy mode	Merging several sources into one destination and append mode are not allowed.
Expiration time	The source value is also copied to the destination.
Replication bandwidth	2 GiB per second per region and per project (physical).

There are separate costs too. Cross-region copy incurs data transfer charges under BigQuery pricing, and you are billed even if you cancel before it finishes. If you copy a snapshot or clone to a new region, a full copy is created, so storage costs rise as well. After the copy, count rows in the destination region to check they match the source, and confirm the destination table permissions are what you intended.

What to do now and what to watch next

Turn on column aliases in a development project first, compare the queries with name conflicts, and then apply it to production. For cross-region copy, check tags and IAM policies first, estimate the cost, and then run bq cp. We also covered two BigQuery updates from the same week: how to use the AI.AGG function and how to use the Gemini button in the SQL editor.