Currently, no. Microsoft is currently researching ways to do this using a combination of work already completed and an estimated execution plan (See the details of their research on the Microsoft Research site ), so we can expect to see something soon. But this is the only development that I am aware of.

The solution I've used with the most success in the past, for processes that take a lot of time, is to break the process up into smaller tasks, and set milestones at the end of each task. The total time for all executions of each task is recorded, and this is used to benchmark the progress of the current execution. This depends heavily on linearity of your queries (i.e. the time taken to execute is linearly proportional to the number of records). Milestones can either be measured by steps through a process, by breaking the data into smaller segments, or both.

The only way you can estimate query execution time by actually running the query. And even after that if may differ next time as it depends on lots of factors like how busy is your server, or many processes are trying to access that table or the quantity of data you are trying to access.

SQL Server doesn't estimate execution time.

Estimated query cost used to be misleadingly equated in the documentation as estimated time in seconds (for features such as query governor ) but it is really just a unitless value.

You can use this below query.

SET SHOWPLAN_ALL ON
GO
SELECT * FROM ENTITY_ATTRIBUTES
GO
SET SHOWPLAN_ALL OFF