Check for existing record before insert

To avoid duplicate record in database, it is better to check for duplicate certain column data before insert. Primary key is a good mechanic to prevent duplicate row but don't count on it. Take a look the following SQL statements.

-- Creating temporary table (Notice the primary key is set to identity)
CREATE TABLE #tblTemp
    ( ID int identity primary key,
           ColA nvarchar(30),
           ColB nvarchar(50)
        );
GO

-- Inserting new record into the temporary table tblTemp
-- If execute this insert statement twice, you will have two row have same data in ColA and ColB
-- with difference ID
INSERT INTO #tblTemp VALUES ('Test A1', 'Test B1');

-- This is better
IF NOT EXISTS (SELECT 1 FROM #tblTemp WHERE #tblTemp.ColA='Test A1')
    INSERT INTO #tblTemp VALUES ('Test A1', 'Test B1')
SELECT * FROM #tblTemp

Another example using sample database AdventureWorks2012.
USE AdventureWorks2012;
GO
SELECT * FROM AdventureWorks2012.Person.Address WHERE AddressLine1='1970 Napa Ct.';
GO

-- We don't want a duplicate record for AddressLine1 '1970 Napa Ct.' and PostalCode 1597
IF NOT EXISTS (SELECT 1 FROM AdventureWorks2012.Person.Address WHERE AddressLine1='1970 Napa Ct.' AND PostalCode=1597)
    INSERT INTO AdventureWorks2012.Person.Address (AddressLine1, City, StateProvinceID, PostalCode)
    VALUES ('1970 Napa Ct.', 'Lane Cove', 50, 1597);

SELECT * FROM AdventureWorks2012.Person.Address WHERE AddressLine1='1970 Napa Ct.';
GO