Thanks svein ...

exactly i was thinking the same
select * from MyTable2 where ID in (<id_we_retrieve_from_the_first_select>)

but i face 2 problems :

1rt: <id_we_retrieve_from_the_first_select> is from 0 to 100, it's mean i will need around 100 statement in my pooling mecanisme (because i need to add "?" as mmuch as i have ids). as i m a little affraid about the multithread implementation of firebird client, every statement have it's own connection and transaction that is not shared. and 100 connections open used ressources ... that not a big deal but a little

2nd: the around time taken for select ... from MyTable2 where ID in (<id_we_retrieve_from_the_first_select>) is 150 ms, that look slow. i reduce from 70 ms by removing the prepare thanks to my pooling mecanisme. but still have 70 ms :( probably i can not do better, it's like this but my first impression is that it's look slow ....

page_reads: 595
page_writes: 4
page_fetches: 2488
page_marks: 4

record_idx_reads: 511
record_seq_reads: 0
record_inserts: 0
record_updates: 0
record_deletes: 0
record_backouts: 0
record_purges: 0
record_expunges: 0