This study investigated, through two experiments, how the design variables of reward programs—namely (1) the evaluation mode of rewards (SE/JE), (2) the temporal distance between reward selection and consumption, and (3) whether a no-choice option is unavailable or available during reward selection decisions—influence customers' preference and choice behavior regarding reward types (hedonic or utilitarian rewards). The research results are as follows. First, Hypothesis 1, which predicted that the relative preference for hedonic rewards over utilitarian rewards would be lower in JE than in SE (evaluation mode effect), was supported for reward types requiring low points but was not supported for reward types requiring high points. Second, Hypothesis 2, which predicted that the relative preference for hedonic rewards over utilitarian rewards would be higher when the reward consumption timing is in the distant future rather than the near future (temporal distance effect), was not supported for reward types requiring low points but was supported for reward types requiring high points. Third, Hypotheses 3 and 4, which predicted that if choice deferral is possible, the evaluation mode effect and temporal distance effect observed when choice deferral is impossible would be weakened or disappear, were supported. Finally, the theoretical and practical implications suggested by the results of this study were proposed, and the limitations of the study and directions for future research were also discussed.