Hi, I have parent-child-child-child-.... (multiple hierarchies). What is the best approach to create tables in data warehouse, when user have no clarity on levels of hierarchies and I need to consider future increase in levels of hierarchies, so I need to plan a structure that can absorb future hierarchies. I saw one example on asktom.oracle.com. Does anybody have better way... select rpad('*',2*level,'*')||ename EmpName, dname, 2 sys_connect_by_path( ename, '/' ) cbp 3 from emp 4 start with mgr is null 5 connect by prior empno = mgr 6 order SIBLINGS by ename 7 / EMPNAME DNAME CBP ------------------------------ -------------- ------------------------------ **KING ACCOUNTING /KING ****BLAKE SALES /KING/BLAKE ******ALLEN SALES /KING/BLAKE/ALLEN ******JAMES SALES /KING/BLAKE/JAMES ******MARTIN SALES /KING/BLAKE/MARTIN ******TURNER SALES /KING/BLAKE/TURNER ******WARD SALES /KING/BLAKE/WARD ****CLARK ACCOUNTING /KING/CLARK ******MILLER ACCOUNTING /KING/CLARK/MILLER ****JONES RESEARCH /KING/JONES ******FORD RESEARCH /KING/JONES/FORD ********SMITH RESEARCH /KING/JONES/FORD/SMITH ******SCOTT RESEARCH /KING/JONES/SCOTT ********ADAMS RESEARCH /KING/JONES/SCOTT/ADAMS Thanks -- //www.freelists.org/webpage/oracle-l