Can HAVING Work Without GROUP BY in SQL Server? (Yes, Here's Why)

Can HAVING work without GROUP BY in SQL Server?

Yes.

And there is a practical reason a data analyst might actually want to do it.

Imagine the finance team asks:

"Alert me only if today's approved refunds exceed $50,000."

You don't need totals by customer, country, product, or payment method.

You need one answer about the entire filtered dataset.

A common instinct is to write:

SELECT
    SUM(RefundAmount) AS TotalApprovedRefunds
FROM dbo.PaymentRefunds
WHERE RefundStatus = 'Approved'
GROUP BY RefundStatus
HAVING SUM(RefundAmount) > 50000;

This works.

But ask yourself:

Why are we grouping by RefundStatus when WHERE already guarantees that every remaining row has RefundStatus = 'Approved'?

We can remove GROUP BY completely:

DECLARE @ReportDate date = '2026-10-05';

SELECT
    SUM(RefundAmount) AS TotalApprovedRefunds,
    COUNT_BIG(*) AS RefundCount
FROM dbo.PaymentRefunds
WHERE RefundStatus = 'Approved'
  AND RefundedAt >= @ReportDate
  AND RefundedAt < DATEADD(day, 1, @ReportDate)
HAVING SUM(RefundAmount) > 50000.00;

What's happening?

Without GROUP BY, SQL Server treats all rows that survive the WHERE clause as one implicit aggregated group.

Think about the query in two stages: