Digitalization has changed existing business models and enabled new ones. This development has been accompanied by the emergence of new pricing options and the possibility of applying established pricing models in new domains. Today, consumers can, for example, pay for accessing a product instead of buying it. Within such sharing services, consumers can usually choose between a flat-rate and a pay-per-use option. Prior work demonstrated that consumers' tariff choices are often systematically biased. Overconfidence was identified as one of the key drivers. Yet, prior research is nonexperimental and focused on the so-called flat-rate bias. By contrast, we examine the effects of overconfidence on tariff choice experimentally. We show that overconfident consumers overestimate their ability to predict their future usage, which leads them to underestimate their actual usage, and eventually leads them to choose a pay-per-use (vs. a flat-rate) option more frequently. We discuss theoretical and managerial implications.