>>Try
>>

>>create table test (Latitude decimal(30,10), Longitude decimal (30,10))
>>
>>declare @CenterLat float=47.7795
>>declare @CenterLon float=-65.7191
>>declare @SearchDistance float=250
>>
>> select * from (select * from test T
>> cross apply (select SIN(@CenterLat/57.2957795130823) * SIN(Latitude/57.2957795130823) + 
>> COS(@CenterLat/57.2957795130823) * COS(Latitude/57.2957795130823) * 
>> COS(Longitude/57.2957795130823 -@CenterLon/57.2957795130823) as Temp) Temp
>>  where (latitude>=@CenterLat - @SearchDistance/111.0 and latitude<=@CenterLat + @SearchDistance/111.0) and
>> (longitude>=@CenterLon - @SearchDistance/111.0 and longitude<=@CenterLon + @SearchDistance/111.0)) X
>> where  3958.75586574 * ACOS(case when Temp > 1 then 1 when Temp < -1 then -1 else Temp end) <=@SearchDistance

>
>This is good code. I also tested it against the main table with close to 2000 records and the related records were returned as is.
>
>Thanks

Can you also test this performance wise using original function approach and using new CROSS APPLY approach and also using Borislav's code?

Just curious for performance.