Let's say I'm making an accounting system. I have Invoices, and on Invoices I have Products. Every invoice can have a set of one or more products, and there could be duplicates. Every product could show up on any number of invoices. The relationship is uni-directional. The Product doesn't need to have a collection of Invoices that contain it. How do I map this? It seems like what I need is a JoinTable specification, but when I create it, I get a table with a unique constraint on invoice_id. That should not be the case; I should be able to use that invoice_id multiple times to be able to specify a set of products. What I need is a join table that would look like this: | invoice_id product_id | 33 58 | 33 37 | 33 962 | but it's failing on the uniqueness constraint of invoice_id. Any ideas on how to specify this? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4016620#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...