A recent study conducted in New York (with Newark serving as control) labeled menus with calorie content, and then questioned patrons whether they noticed the labels, and if they did: did they actually use them to make better choices?
As it turns out: not so much:

Those people that did notice the menu labels used them…to buy more calories. Not significant at this point, but disheartening to say the least. Sodium intake also increased.
I guess the answer is to label menus… but lie.
Ummm, huzzah for science!