Two experiments investigated the impact of warnings in mass-media advertising in print (magazine) or broadcast (television) alcoholic beverage ads. Experiment 1 showed that participants exposed to highly conspicuous warnings in print ads performed better on subsequent memory and knowledge tests than those exposed to less conspicuous warnings. Less conspicuous warnings were generally no better than no warnings. In experiment 2, participants viewing broadcast commercials with both-modality warnings generally performed better than those exposed to voice-only or no warnings. Print-only warnings produced performance equivalent to, or in one case lower than, both-modality warnings. Both experiments show that warnings in ads can communicate information if presented in a salient form.