To investigate risk behavior, behavioral economic researchers use gambles consisting of many trials. But participants usually do not receive the sum of all outcomes. Instead, participants are paid out according to one trial that is randomly selected in the end. As the payment method might influence risk behavior, in this study we let participants play a gamble two times. In the pay-all condition, the outcome of every trial is paid out, and in the pay-one condition, the outcome of one randomly selected trial, multiplied by the number of trials, is paid out. Our results show that participants made 10 % more risky decisions in the pay-all compared to the pay-one condition. In an additional questionnaire with hypothetical risk choices, participants also showed substantially riskier behavior in the pay-all condition. Risk behavior was highly correlated in both conditions.The results suggest that the absolute level of risk choices is higher in pay-all paradigms than in payone paradigms.