the stack trace says it all.... there is no way for hibernate to figure out the unique rows from a joined resultset with multiple bags. Thus instead of creating duplicate entries in your llist we fail with the stacktrace. In the future we plan to provide a fallback for this scenario where a second (sub)select will be done to load the other bags. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3986847#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...