Much work on the public health communication component of social marketing focuses on message development. But there is good evidence that failure and success in public health communication is better predicted by variation in exposure to messages achieved than it is by variation in quality of messages. The inconsistent results about effects from some major projects (Stanford Heart Disease, Minnesota Heart Health, Pawtucket Heart Health, COMMIT) may reflect their lack of success in obtaining heavy exposure to their messages. Those results contrast with the successful results of a variety of other programs, particularly kitchen sink programs, which have been able to obtain higher levels of exposure and have some evidence of important effects.