必须同时使用count（*）和group by：

select count (*),DuplicateField from table
group by  DuplicateField
having count (*)>1

因此，您将获得记录。如果你想删除，我可以举个例子告诉你：

select 
    'RepetedRows',
    SA.Saleref,
    MIN(CAST(SA.SaleAdditionID  as varchar(100)) ),
    'DELETE sms.tblsaleadditions    WHERE SaleAdditionID = ''' + MIN(CAST(SA.SaleAdditionID as varchar(100)) ) + ''''
    --select * 
from 
    sms.tblsaleadditions    SA
inner join (
    select
        saleref,
        additionref,
        count(*)    N,
        SUM(AdditionAmount) nn
    from 
        sms.tblsaleadditions
    group by 
        saleref,
        additionref
    having count(*) > 1
    ) V ON SA.saleref = V.Saleref and SA.additionref = V.additionref 
group by 
    SA.Saleref

现在您可以复制所有delete语句并运行它们。

在mysql查询中这是什么？我们有这个要求，但是现有的数据库是mysql。

让我假设你有另一个专栏，比如 col3 --可以区分重复行的。然后你可以做：

delete t
    from t t join
         (select cfield1, cfield2, min(col3) as min_col3
          from t
          group by cfield1, cfield2
         ) tt
         using (cfield1, cfield2)
    where t.col3 > tt.min_col3;

如果没有这样的列，那么我建议创建一个新表。下面使用了mysql的一个特性，这是一个非常不鼓励使用的特性，但是它应该做你想做的事情：

create table temp_t as
    select t.*
    from t
    group by col1, col2;

truncate table t;  -- be very careful here, copy the data somewhere else first!

insert into t
    select *
    from temp_t;