No. &(expr) is not an lvalue. Lvalues are from the set of identifer primary[expression] lvalue.identifier primary->identifier *expression (lvalue) You will note that "& thingy" is not one of them. So you are syntactically incorrect. -Ron