A common assumption underlying the analysis of consumers' choices among optional tariffs is that consumers choose the tariff that maximizes their surplus and, thus, the tariff that leads to the lowest billing rate for a given amount of usage.Yet there is evidence that many users prefer a flat rate even though their billing rate would be lower with a pay-per-use tariff (flat-rate bias), and some users prefer a pay-per-use tariff even though they would save money with a flat rate (pay-per-use bias). The authors conduct four empirical analyses based on three different data sets. They show that the flat-rate bias is more important and has a greater regularity and time persistence than the pay-per-use bias. They classify potential causes of the flat-rate bias as “insurance effect," “taxi meter effect," “convenience effect," and “overestimation effect" and show that the insurance, the taxi meter, and the overestimation effects lead to a flat-rate bias. They provide evidence that underestimation of usage is a major cause of the pay-per-use bias. They show that the flat-rate bias does not significantly increase customer churn and thus results in a short- and long-term profit increase. In contrast, the pay-per-use bias largely increases churn so that in the long run, the additional short-term profit is offset by higher churn.