hi, I'm having trouble constructing a hierarchical query. Can someone help. SQL> select * from hier; A B ---------- ---------- 99 7 100 1 100 2 1 2 2 3 4 8 200 6 7 rows selected. The order of the output is not important. The output can be: 100 1 1 2 2 3 100 2 2 3 -> should not display this duplicate 200 6 OR 100 1 100 2 1 2 2 3 2 3 -> should not display this duplicate 200 6 We will start with a number. In this case, 100. If a=100, get the corresponding value of b and match that back to a and get the next corresponding value of b until b!=a. After the first iteration, now get the next value of a>=100 and process it in the same manner. Duplicate records should only be displayed once, in this case, 2 3 should only be displayed once. Is it possible to achieve the requirements in a single query? thanks. susan -- //www.freelists.org/webpage/oracle-l