there are three ways-
1. select distinct records and insert it into another dummy table, then truncate all records from original table and then copy records form dummy table to the original table.
2. add Identity column to the table for time being which will make the records distict and based on values of rest of the columns you can delete the duplicate records- search for this on google. you will get it.
3. If you are using SQL Server 2005/8 you can use ROW_NUMBER function to delete the duplicate records.