Delay discounting is the decline in the present value of a reward with delay to its receipt. (Mazur,1987). The delay discounting task is used to measure delay discounting rate, which requires the participants to choose between two options: one involves immediate delivery of a reward, and other involves delivery after a delay, and the immediate rewards are adjusted in value until the subject feels there is no difference between the immediate and the delayed reward. Some previous studies (Robles and Vargas, 2007; 2008; Robles et al., 2009) found that the order of presentation of the immediate rewards (ascending or descending) significantly influenced the estimated delay discounting rate, which is known as the order effect. Uncertainty about the future and impulsivity could explain delay discounting behavior. The purpose of this study was to explore the order effect in delay discounting assessment. The current study found that the order effect in the delay discounting task can be explained by risk aversion, loss aversion and impulsivity. In the current study, the two kinds of fixed procedure (ascending and descending), and the titrating delay discounting task were used to estimate the degree of delay discounting. Also, two gambling tasks were applied to measure risk and loss aversion indices. The BIS-11 scale was used to assess the level of trait impulsivity. The results indicated that impulsivity biases individuals to choose the immediate small reward rather than the large delayed reward, resulting in lower area under the discounting curve (AUC) when estimated with the ascending-sequence delay discounting task. Also, impulsivity moderated the relationship between loss aversion and AUC estimated with the descending-sequence delay discounting task.