470

I know I've done this before years ago, but I can't remember the syntax, and I can't find it anywhere due to pulling up tons of help docs and articles about "bulk imports".

Here's what I want to do, but the syntax is not exactly right... please, someone who has done this before, help me out :)

INSERT INTO dbo.MyTable (ID, Name)
VALUES (123, 'Timmy'),
    (124, 'Jonny'),
    (125, 'Sally')

I know that this is close to the right syntax. I might need the word "BULK" in there, or something, I can't remember. Any idea?

I need this for a SQL Server 2005 database. I've tried this code, to no avail:

DECLARE @blah TABLE
(
    ID INT NOT NULL PRIMARY KEY,
    Name VARCHAR(100) NOT NULL
)

INSERT INTO @blah (ID, Name)
    VALUES (123, 'Timmy')
    VALUES (124, 'Jonny')
    VALUES (125, 'Sally')

SELECT * FROM @blah

I'm getting Incorrect syntax near the keyword 'VALUES'.


  • Your above code is fine just need to add ',' after values statement - sam
  • INSERT INTO @blah (ID, Name), VALUES (123, 'Timmy'), VALUES (124, 'Jonny'), VALUES (125, 'Sally') - sam
  • Just a caution: you can insert up to 1000 rows only by this method. INSERT INTO #Test (LWPurchaseOrderID )VALUES ( 935791 ), ( 935933 ) - Anoop Verma
  • 2005 is no longer supported. For 2008, 2012 and 2016 you can almost use what you put INSERT INTO @blah (ID, Name) VALUES (123, 'Timmy'), (124, 'Jonny'), (125, 'Sally') "VALUES" only appears once and you need commas between the sets. - J. Chris Compton

13 답변


297

INSERT INTO dbo.MyTable (ID, Name)
SELECT 123, 'Timmy'
UNION ALL
SELECT 124, 'Jonny'
UNION ALL
SELECT 125, 'Sally'

For SQL Server 2008, can do it in one VALUES clause exactly as per the statement in your question (you just need to add a comma to separate each values statement)...


  • Is this any more efficient than using multiple INSERT statements? - Code Commander
  • @Code Commander: no, in that it is longer to compile. Yes, in that you have one insert only. But it answers the question: no repeat of the INSERT table (columnlist) - gbn
  • @VoidKing I know this comes half a year later and you might have figured this out yet long ago, but it's really quite simple. By using select you create a set with columns and rows, and by design these rows can be inserted into another table with an equal amount of columns. You can even use a mixture of literals and values. For example, using insert with select 'A', ID from ATable would insert 'A' in the first column every time and the ID column value of the corresponding row of ATable in the second column. - MarioDS
  • This also workes with DB2 which is an important sidenote for those of us stuck in outdated technology. The values seperated by comma answer is better in my mind for those of you working in SQL Server 2008 or newer. The OP can remove all "values" except the first and replace with a , - JPK
  • @PRMan you would not do that after version SQL Server 2008. As mentioned... - gbn

457

Your syntax almost works in SQL Server 2008 (but not in SQL Server 20051):

CREATE TABLE MyTable (id int, name char(10));

INSERT INTO MyTable (id, name) VALUES (1, 'Bob'), (2, 'Peter'), (3, 'Joe');

SELECT * FROM MyTable;

id |  name
---+---------
1  |  Bob       
2  |  Peter     
3  |  Joe       

1 When the question was answered, it was not made evident that the question was referring to SQL Server 2005. I am leaving this answer here, since I believe it is still relevant.


  • But don't in SQL Server 2005. msdn.microsoft.com/en-us/library/ms174335.aspx: SQL Server 2008 introduces the Transact-SQL row constructor (also called a table value constructor) to specify multiple rows in a single INSERT statement. - abatishchev
  • @abatishchev: Yes, but when answer the question (4 minutes after it was posted), the question was not tagged 'sql-server-2005'. In fact the OP later edited the post to specify this, and it is clear from the 'EDIT' label. The original question was simply tagged 'tsql'. - Daniel Vassallo
  • Server 2008 doesn't allow more than 1000 rows inserted this way. - Michael
  • What happens, if one value set is faulty? Will all inserts rolled back or just the faulty row? - netblognet
  • @netblognet I just tested that only faulty rows are not inserted (all of the all the others are inserted correctly) - Mauricio Gracia Gutierrez

216

If your data is already in your database you can do:

INSERT INTO MyTable(ID, Name)
SELECT ID, NAME FROM OtherTable

If you need to hard code the data then SQL 2008 and later versions let you do the following...

INSERT INTO MyTable (Name, ID)
VALUES ('First',1),
('Second',2),
('Third',3),
('Fourth',4),
('Fifth',5)


16

Using INSERT INTO ... VALUES syntax like in Daniel Vassallo's answer there is one annoying limitation:

From MSDN

The maximum number of rows that can be constructed by inserting rows directly in the VALUES list is 1000

The easiest way to omit this limitation is to use derived table like:

INSERT INTO dbo.Mytable(ID, Name)
SELECT ID, Name 
FROM (
   VALUES (1, 'a'),
          (2, 'b'),
          --...
          -- more than 1000 rows
)sub (ID, Name);

LiveDemo


This will work starting from SQL Server 2008+


  • Can I have a link to an article about this 'sub' syntax. - CodeCamper
  • @CodeCamper docs.microsoft.com/en-us/sql/t-sql/queries/… section: C. Specifying multiple values as a derived table in a FROM clause - Lukasz Szozda
  • The advantage of this answer is that it provides a way to specify identical values without repeating them (which is what I was looking for). E.g. with a third column that is identical, you wouldn't need to repeat it for a thousand times. - Vadim Berman
  • @VadimBerman Yes, that is good scenario when there is no default defined on table. - Lukasz Szozda

13

You could do this (ugly but it works):

INSERT INTO dbo.MyTable (ID, Name) 
select * from
(
 select 123, 'Timmy'
  union all
 select 124, 'Jonny' 
  union all
 select 125, 'Sally'
 ...
) x


8

You can use a union:

INSERT INTO dbo.MyTable (ID, Name) 
SELECT ID, Name FROM (
    SELECT 123, 'Timmy'
    UNION ALL
    SELECT 124, 'Jonny'
    UNION ALL
    SELECT 125, 'Sally'
) AS X (ID, Name)


6

This looks OK for SQL Server 2008. For SS2005 & earlier, you need to repeat the VALUES statement.

INSERT INTO dbo.MyTable (ID, Name)  
VALUES (123, 'Timmy')  
VALUES (124, 'Jonny')   
VALUES (125, 'Sally')  

EDIT:: My bad. You have to repeat the 'INSERT INTO' for each row in SS2005.

INSERT INTO dbo.MyTable (ID, Name)  
VALUES (123, 'Timmy')  
INSERT INTO dbo.MyTable (ID, Name)  
VALUES (124, 'Jonny')   
INSERT INTO dbo.MyTable (ID, Name)  
VALUES (125, 'Sally')  


6

It would be easier to use XML in SQL Server to insert multiple rows otherwise it becomes very tedious.

View full article with code explanations here http://www.cyberminds.co.uk/blog/articles/how-to-insert-multiple-rows-in-sql-server.aspx

Copy the following code into sql server to view a sample.

declare @test nvarchar(max)

set @test = '<topic><dialog id="1" answerId="41">
        <comment>comment 1</comment>
        </dialog>
    <dialog id="2" answerId="42" >
    <comment>comment 2</comment>
        </dialog>
    <dialog id="3" answerId="43" >
    <comment>comment 3</comment>
        </dialog>
    </topic>'

declare @testxml xml
set @testxml = cast(@test as xml)
declare @answerTemp Table(dialogid int, answerid int, comment varchar(1000))

insert @answerTemp
SELECT  ParamValues.ID.value('@id','int') ,
ParamValues.ID.value('@answerId','int') ,
ParamValues.ID.value('(comment)[1]','VARCHAR(1000)')
FROM @testxml.nodes('topic/dialog') as ParamValues(ID)


6

USE YourDB
GO
INSERT INTO MyTable (FirstCol, SecondCol)
SELECT 'First' ,1
UNION ALL
SELECT 'Second' ,2
UNION ALL
SELECT 'Third' ,3
UNION ALL
SELECT 'Fourth' ,4
UNION ALL
SELECT 'Fifth' ,5
GO

OR YOU CAN USE ANOTHER WAY

INSERT INTO MyTable (FirstCol, SecondCol)
VALUES 
('First',1),
('Second',2),
('Third',3),
('Fourth',4),
('Fifth',5)


6

I've been using the following:

INSERT INTO [TableName] (ID, Name)
values (NEWID(), NEWID())
GO 10

It will add ten rows with unique GUIDs for ID and Name.

Note: do not end the last line (GO 10) with ';' because it will throw error: A fatal scripting error occurred. Incorrect syntax was encountered while parsing GO.


5

Corresponding to INSERT (Transact-SQL) (SQL Server 2005) you can't omit INSERT INTO dbo.Blah and have to specify it every time or use another syntax/approach,


2

This is working very fast,and efficient in SQL. Suppose you have Table Sample with 4 column a,b,c,d where a,b,d are int and c column is Varchar(50).

CREATE TABLE [dbo].[Sample](
[a] [int] NULL,
[b] [int] NULL,
[c] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[D] [int] NULL
)

So you cant inset multiple records in this table using following query without repeating insert statement,

DECLARE @LIST VARCHAR(MAX)
SET @LIST='SELECT 1, 1, ''Charan Ghate'',11
     SELECT 2,2, ''Mahesh More'',12
     SELECT 3,3,''Mahesh Nikam'',13
     SELECT 4,4, ''Jay Kadam'',14'
INSERT SAMPLE (a, b, c,d) EXEC(@LIST)

Also With C# using SqlBulkCopy bulkcopy = new SqlBulkCopy(con)

You can insert 10 rows at a time

   DataTable dt = new DataTable();
        dt.Columns.Add("a");
        dt.Columns.Add("b");
        dt.Columns.Add("c");
        dt.Columns.Add("d");
        for (int i = 0; i < 10; i++)
        {
            DataRow dr = dt.NewRow();
            dr["a"] = 1;
            dr["b"] = 2;
            dr["c"] = "Charan";
            dr["d"] = 4;
            dt.Rows.Add(dr);
        }
        SqlConnection con = new SqlConnection("Connection String");
        using (SqlBulkCopy bulkcopy = new SqlBulkCopy(con))
        {
            con.Open();
            bulkcopy.DestinationTableName = "Sample";
            bulkcopy.WriteToServer(dt);
            con.Close();
        }


1

This will achieve what you're asking about:

INSERT INTO table1 (ID, Name)
    VALUES (123, 'Timmy'), 
           (124, 'Jonny'), 
           (125, 'Sally');

For future developers, you can also insert from another table:

INSERT INTO table1 (ID, Name)
    SELECT 
         ID, 
         Name 
    FROM table2

Or even from multiple tables:

INSERT INTO table1 (column2, column3)
    SELECT 
         t2.column, 
         t3.column
    FROM table2 t2
         INNER JOIN table3 t3
         ON t2.ID = t3.ID

Linked


Related

Latest