Ive read the books online and the pages on this great site. I am now trying to determine why my server's performance is horrible. I am now trying to take theory (what i have read) and translate it into reality. The index tuning wizard improved my databse's speed, but a few seem questionable and I would like to i understand it. this:http://www.sql-server-performance.com/clustered_indexes.asp suggests both avoid redundant indexes and avoid multi-column clustered indexes. Which should be by clustered index (all fields used often) CREATE CLUSTERED INDEX [IX_Host1] ON [dbo].[Host]([HostId]) OR CREATE CLUSTERED INDEX [IX_Host1] ON [dbo].[Host]([HostId], [source], [field3], [field4] ) The tuning wizard created the single column index as clustered and the second index as non-clustered. Can anyone explain why this is better? Context for the question: My DB has a lot of writes and few reads. I has a table, called host, that contains information on computers. This table is refrenced often, most often based on the 'hostid', a sequential identifier for the host. Since several systems generate the data used in this DB it is not truely sequential on its' own. There are multiple ID '1', multiple ID '2', and so on. The hostid paired with the 'source' creates a unique refrence. Because of this EVERY refrence to the host table will pair the required information with the 'source' field. -- Tom Motivation: If pretty poster and a cute saying are all that it takes to motivate you, you probably have a very easy job. The kind robots will be doing soon.