It is generally held that repetition enhances the effectiveness of television advertisement in terms of information retention and affective rating. The present experiment examined this effect in light of similar predictions made by the encoding variability and mere exposure hypotheses. Seven different groups of 17 subjects each served in a between groups design. Conditions were defined by different levels of presentation frequency (one, three, and six) and repetition similarity (low, medium, and high). The results suggest that greater frequency and lower similarity (higher variability) of repetition enhance retention measures. No firm conclusions could be made concerning the affective ratings. These findings were interpreted as providing support for the encoding variability hypothesis, but not for the mere exposure hypothesis. The implication of these findings for advertising repetition, in particular, and stimulus repetition, in general, were discussed.