LESSCHARSET=latin1 but still not latin1 in less?

[y551025]

Hello --
I'm having trouble seeing certain special characters with less.
Here's my situation:
% less --version
less 382+iso254
% printenv | grep -i lesschar
LESSCHARSET=latin1

(and I'm running SuSE9.3).

When I try to view a file, foobar, containing latin1 text:
% less foobar
I find that things like u-with-umlaut and e-with-acute appear as space
characters (and the following character is missing, too).
So it looks like the  less  program is not "getting my message" about
character set.

Any suggestions welcome.
Thanks.