Consumers' search for information on the Internet is predicted to become the most popular activity associated with commerce on the Internet. Recent studies have revealed a growing number of car buyers who research their automotive purchases on the Internet. However, little knowledge about Internet advertising related to online consumer information search has been accumulated in the area of Internet advertising research. To date, we have rarely seen studies, specifically experimental ones, of attitudes toward Internet advertising using attitude function theory, although this is a recently re-highlighted research approach to persuasive communication. Hence, the main purpose of this dissertation study was to advance the understanding of consumers' attitudes toward Internet advertising as persuasive communication by the extension and the application of attitude function theory to the new medium. On the basis of a comprehensive literature review, the analysis of recent research on Internet advertising as persuasive communication identified a knowledge gap in Internet advertising research: Previous studies have virtually ignored the effects of advertising messages in terms of consumers' psychological needs and associated attitudes in order to understand persuasive communication effects on the Internet. Attitude function theorists predict that the effectiveness of persuasive communication messages will be greater when the messages match attitude functions of message recipients than when the messages do not match such attitude functions. Thus, applying attitude function theory to Internet advertising research area, this study tested functional matching hypotheses in relation to a new media context such as the Internet. Following a growing, alternative approach to the study of Internet advertising, this dissertation examined the persuasive communication effects of commercial websites as a format of Internet advertising. A focus group discussion and two pilot studies were conducted as the preliminary research for this dissertation. Two main experiments were performed to test research hypotheses and answer the research question for this study. Attitude function-based websites related to car ix information search were constructed as the stimulus materials for the main experiments. Experiment 1 tested functional matching hypotheses. Experiment 2 replicated Experiment 1 and also manipulated corporate credibility as a special case of source credibility in a form of publicity. Attitude function was measured via the attention-to-social-comparison-information (ATSCI) scale as the revised self-monitoring scale in both experiments. Multivariate analysis of variance (MANOVA) tests were conducted for the statistical analysis of data, mainly because previous traditional and Internet advertising research suggested empirical evidence that attitude toward the site, attitude toward the brand, and behavioral intention to revisit the site as the dependent variables in this study are significantly correlated. The results of the MANOVA test for Experiment 1 did not find empirical evidence suggesting a significant functional matching effect on the combination of three dependent variables. On the other hand, the MANOVA results for Experiment 2 found empirical evidence that there was a significant multivariate main effect of corporate credibility on the combination of three dependent variables even though there were no significant interaction effects among attitude function (as measured by the ATSCI scale), attitude function-based website appeals, and corporate credibility on the combination of the dependent variables. Discussion of results provided theoretical explanations for inconsistent findings from the experiments. Limitations of this study are discussed, and several suggestions for future research are provided.