Using telephone and Internet (web-based) survey samples for a national advisory referendum, this study investigates whether the underlying preferences across two survey modes are equivalent. Results from a structural test support the conclusion of common preferences.