This study assessed the accuracy of judgments of 100 school-age children (5-11 years of age) as to the presence of real fruit content in three sets of cereals and beverages advertised on TV: real fruit, nonfruit, and artificially flavored products. In the baseline session accuracy increased with age, but children at each age misjudged the real fruit content of artificial fruit products. In the naturalistic viewing session children judged fruit content of 12 advertised products, six of which were embedded in a program and six for which advertisements were not shown. Control group children saw toy advertisements on TV, and then judged the 12 food products. Subsequently, children in both groups saw the same advertisements they had seen earlier, this time without the program and with instructions to attend to the messages carefully. Few significant changes were found between the baseline and naturalistic viewing sessions. Following instructions to attend carefully, accuracy of judgment of advertised artificial fruit products was lower than that of the baseline session among experimental group children. By contrast, accuracy, under instructions to attend carefully, was higher than in the baseline session both for control children's judgments of "advertised" products and for experimental children's judgments of products not advertised. (Author/SS)