SELECT i.* from table1 i
   WHERE (( i.field2 = "FA" and i.field3 = 0)
	   OR EXISTS (SELECT * from table2 s WHERE s.key1 = i.key1 and s.fielda <> "X" and s.fieldb <> "I"))
	   AND NOT EXISTS(SELECT * from table3 a WHERE a.key1 = i.key1 AND a.field4 = i.field4)
   ORDER BY i.field4

>I need to do the following:
>
>

>SELECT i.* from table1 i
>   WHERE (IF exists) (SELECT * from table2 s WHERE s.key1 = i.key1 and s.fielda <> "X" and
>                      s.fieldb <> "I")
>   AND i.field2 = "FA" and i.field3 = 0
>   AND NOT EXISTS(SELECT * from table3 a WHERE a.key1 = i.key1 AND a.field4 = i.field4)
>   ORDER BY i.field4
>

>In other words,
>
>1. If a record from table1 is in table2, I want to include records from table1 if table2's fielda <> "X" and table2's fieldb <> "I", but if records in table1 whose field2 = "FA" and field3 = 0 are NOT in table2, I want to include those as well.
>2. Of those records in the result from #1, I only want to include the records that are NOT in table3 (determined by key1 and field4 field matches). (I want to strip out any records from #1 that are already in table3.
>
>Right now I am using:
>

>SELECT i.* from table1 i
>   WHERE EXISTS (SELECT * from table2 s WHERE s.key1 = i.key1
>                 AND s.fielda <> "X" AND s.fieldb <> "I")
>   AND i.field2 = "FA" and i.field3 = 0
>   AND NOT EXISTS(SELECT * from table3 a WHERE a.key1 = i.key1
>                  AND a.field4 = i.field4)
>   ORDER BY i.field4
>

>
>But that leaves out those records from table1 that are NOT in table2 which I want to include in the result if they are not in table2 and are not in table3 (if table1.field2 = "FA" and table1.field3 = 0)...