HELP: Query Needing tuned

Hi there,

I've recently been tasked with getting some legacy SQL reports working at a decent speed on our new SQL Server 2008 box. These reports used to take 18+ hours to run and for the most part I've been successfull in greatly improving their performance! However I'm having particular trouble with a small and relatively simple LOOKING query.... I should also note that I don't properly understand what the writer was trying to achieve with this query, but it is used several times throughout several reports (each time taking about 2 minutes to run which adds up to a very significant slow down)

Code:

declare @xdate as Datetime
set @xdate = CONVERT(DATETIME, '31 Dec 3000')

SELECT
    A.IDField,
    NEXT_RETURN_DATE =
                    MIN(CASE
                        WHEN A2.EndDate BETWEEN DATEADD(S, 1, A.EndDate) AND DATEADD(M, 3, A.EndDate) THEN A2.EndDate
                        ELSE @xdate
                    END),
    NEXT_START_DATE =
                MIN(CASE
                    WHEN A2.EndDate BETWEEN DATEADD(S, 1, A.EndDate) AND DATEADD(M, 3, A.EndDate) THEN A2.StartDate
                    ELSE @xdate
                END)
FROM Table1 A
LEFT JOIN Table1 A2 ON A.NonIDField1 = A2.NonIDField1 AND A.NonIDField2 = A2.NonIDField2
WHERE  A.EndDate BETWEEN '01 Jan 2010' AND '01 Feb 2011 23:59'
GROUP  BY A.IDField

I'd love to hear some suggestions on how I could improve this or why it's taking so long to execute.

Looking at the execution plan doesn't seem to help me any in this particular case:

Select 0% <- Hash Match (Aggrigate) 5% <- Hash Match (left outer join) 15% <- A & B

A <- Compute Scalar 2% <- Table scan Table 1 39%
B <- Table Scan Table2 39%

Table has about 90k rows being used for this query! Hardly anything....

Thanks, and sorry if this is a stupid question! I'm new here but looking forward to learning and hopefully eventually helping some people out myself!

- Angus