--TA表 表中数据有160万条记录
   create table TA
   (BH VARCHAR(10) NULL,
    HW VARCHAR(10) NULL,
    SL INT
    )
 --TB表 表中数据有10万条记录
  CREATE TABLE TB
   (RQ VARCHAR(10 NULL),
    BH VARCHAR(10) NULL,
    HW VARCHAR(10) NULL,
    SL INT
   )

SELECT BH,HW,SL FROM TA WHERE
not exists (select 1 from TB where TB.BH+TB.HW=TA.BH+TA.BH);

------解决方案--------------------
bh+hw 是用不到索引的
改成下面的会好些，表a b 上面要有索引。
SQL code
SELECT BH,HW,SL FROM TA WHERE
not exists (select 1 from TB where TB.BH=TB.HW and TA.BH=TA.BH);

------解决方案--------------------
SQL code
select BH,HW,SL from TA
where not exists (select 1 from TB where TB.BH+TB.WH=TA.BH+TA.WH)

------解决方案--------------------
数据量这么大，怀疑SQL SERVER根本不会用到索引。 基本照着HASH去了，可以贴一下执行计划吗？

------解决方案--------------------
SELECT TA.BH,TA.HW,TA.SL FROM TA,TB WHERE
CHARINDEX(TA.BH+TA.BH,TB.BH+TB.HW)<=0

------解决方案--------------------
SQL code


--if object_id('TA') is not null 
--    drop table TA
--Go
--create table TA
--(
--BH VARCHAR(10) NULL,
--HW VARCHAR(10) NULL,
--SL INT
--)
--if object_id('TB') is not null 
--    drop table TB
--Go
--CREATE TABLE TB
--(
--RQ VARCHAR(10) NULL,
--BH VARCHAR(10) NULL,
--HW VARCHAR(10) NULL,
--SL INT
--)


IF OBJECT_ID('tempdb..#tmp_TA') IS NOT NULL
    DROP TABLE #tmp_TA
SELECT *,BH+HW AS IDEN INTO #tmp_TA FROM TA

IF OBJECT_ID('tempdb..#tmp_TB') IS NOT NULL
    DROP TABLE #tmp_TB
SELECT *,BH+HW AS IDEN INTO #tmp_TB FROM TB


CREATE NONCLUSTERED INDEX #tmp_TA_IDEN ON #tmp_TA(IDEN)
CREATE NONCLUSTERED INDEX #tmp_TB_IDEN ON #tmp_TB(IDEN)


SELECT BH,HW,SL FROM #tmp_TA AS a WHERE NOT EXISTS(SELECT 1 FROM #tmp_TB AS x WHERE a.IDEN=x.IDEN)