Abstract : Abstract Thanks to an experimentation, this research studies the effects of two broadcast sponsorships on implicit attitude toward a new brand. This brand sponsors two quasi-identical broadcasts differentiated only by their emotional content. To be ecologically valid, the experiment records the attitudinal effects one week later when the subjects don't remember the exposition phase any more and cannot explicitly recognize the brand. The attitude is then implicit. The impact of humorous broadcast sponsorship (with two exposures) on implicit attitude toward the brand is manifest and positive. The sad broadcast has no implicit effects: this type of TV sponsorship has been inefficient. The results are explained with the familiarity-misattribution model. At least, practical suggestions are proposed.