-N will export char data as Unicode

so use both -W and -N switch? also, what would i see different in the file to confirm its in unicode?

The file size should double, for example, for a unicode file. (2 bytes per character.)

A text editor such as Ultraedit would confirm (Ctrl-H)

ok Thanks...So, do I use both -w and -n ?

Something like this will produce Unicode output.

bcp "select N'test' as test" queryout test.txt -w -T

You can check the format of the output file produced by opening it in Notepad.exe, then choose File...Save As... and at the bottom of the dialog window will be the file type Unicode.

I generated the file with w option and it does look different than the file without w option. But I need to make sure if it is in Unicode format. I tried to open it in notepad then save as, The default encoding type is ANSI.

Also opened in notepad++, it says ANSI.

Does your notepad++ have a hex editor?

In Ultraedit I can see the hex value of the bytes, and I can clearly see in a unicode file that there are 2 bytes for each character. (I click on a character and 2 bytes are highlighted.)

cast your column as unicode

David, I downloaded ultraedit. Is unicode file always shows in hexadecimal? I did what you suggested and when I click on 1 character, two bytes are highlighted. But when I opened a non-unicode file, it doesn't open up in hexadecimal.

I am using the following swtich in BCP.

-w -T -t"|" -r"" -c -q

Seems like one switch is overiding the other. File looks different but enocoding hasn't changed to unicode.