Research Question While product returns (PR) are regarded as an unwanted, but integral element in the firm-customer exchange process in ecommerce, they are associated with a series of implications and consequences. First, PR diminish net sales and are costly. Second, they potentially harm the environment. Third, PR imply a series of process steps and thus effort and hassle for the customers. Albeit these diverse consequences, existing studies have-to our knowledge-so far neglected the influence of customers' evaluations on PR on their PR-behavior. In this study, we put customers' views and evaluations about PR into the center of our attention. We furthermore explore the underlying beliefs based on which these views and evaluations are formed. We therefore draw on Ajzen's theory of planned behavior and attempt to answer the following research questions: 1. How do customers' intentions to return products after online shopping affect actual rates of product return in ecommerce? 2. How do customers' attitudes, subjective norms and perceived behavioral control affect their intention to return products after online shopping? 3. What are the underlying beliefs which lead to the formation of customers' attitudes, subjective norms and perceived behavioral control towards product returns? Method and Data In order to test and refine the measurement instruments of our main study and to freely elicit salient customer beliefs, we conducted a pilot study with 58 participants. We therefore used open questions for the belief elicitation part and analyzed the qualitative inputs with regard to content and frequency, thus identifying the most frequent beliefs. We then included two items for each belief in our main study to assess the validity of the beliefs elicited. Our main study was conducted in cooperation with a large European online fashion retailer by surveying customers in France, Germany, Sweden and Italy who had been active within a 3-months timeframe before. Since a unique customer- ID was used for each of the 338 participants, we are able to match the questionnaire data with actual transaction-data. We use established scales for our survey and apply commonly accepted approaches to test for potential common-method, non-response and informant bias. In order to assess model fit and to test for hypothesized effects, we conducted a confirmatory factor analysis and used structural equation modelling in AMOS Graphics 23. We furthermore use a MIMIC-model to assess the influence of the beliefs elicited before. Summary of Findings Our initial results suggest customers' intention to return to influence their actual PR-behavior. Furthermore, customers' intentions to return are influenced by their attitudes and subjective norms (attitudes of important others). With regard to the underlying beliefs that lead to the formation of attitudes, subjective norms and perceived behavioral control, we find different sets of beliefs to be dominant. First, instrumental or hedonic behavioral beliefs associated with the facilitating role of PR ("reduced risk of online shopping" and "enabler to forgo shopping trips to local stores") are found to significantly influence the formation of attitudes towards PR. Surprisingly, other salient-but rather negative-beliefs such as "potential environmental damage," "reduced consumer awareness" or "economic losses for local retailers" are not found to be significant in this context. Furthermore, customers regard their online retailer and their parents as important, when forming subjective norms. Third, "return stickers" and "reusable packaging material" delivered with the package as well as an easy PR-process are regarded as important, when forming PBC. Key Contributions With the present study, we expand of the research field on drivers of customer product return behavior, helping future research to build more adequate research models and hypotheses. We adapt Ajzen's Theory of Planned Behavior to a new field of application and give further confirmation to it, as we find significant relationships using real data on actual behavior, while many other studies (on other types of behavior) only have access to self-reported data. Based on the results of our study, practitioners might consider to mildly influence the image of PR (e.g. by mildly educating customers regarding the negative side-effects of PR), thus influencing customers' attitudes and subjective norms. They can design communication measures to influence the views and evaluations of customers towards PR in a more targeted way, as our study provides evidence for the underlying beliefs that lead to the formation of their views and evaluations. [ABSTRACT FROM AUTHOR]