Decision making under risk changes with age. This age-related difference is often attributed to the alteration in risk tolerance. A recent age-related decision making meta-analysis suggested that in decisions from experience, age-related differences in risk taking were a function of decreased learning performance, whereas in decisions from description, younger adults and older adults showed similar risk-taking behavior. However, little is known about whether age-related differences in decision making reflect a direct effect of aging or, alternatively, caused by age-related changes in risk attitudes, like impulsiveness. Our main goal was to investigate older adults’ risk propensity in different decision making tasks, namely, description-based and experience-based tasks. Moreover, this study was to determine whether self-report of impulsiveness could predict behavior on different decision tasks, and whether this prediction could be mediated by age. We tested two groups of participants: 27 older adults (age 65.56±5.38; 29.63% male) and 27 younger adults (age 21.04±1.60; 40.74% male). All participants completed a battery of psychometric tests including Barratt Impulsiveness Scale-11th version (BIS-11), and two sets of decision making tasks: a description-based task ? Cambridge Gambling Task (CGT) and an experience-based task ? Balloon Analogue Risk Task (BART). During the CGT, participants were told that a token was hidden under one of the red or blue cards, with color ratios changing from 5:1, 4:2, to 3:3. To indicate the location of the token, participants had to bet on one color of the cards, and the bets associated with a certain amount ratio (10 vs. 90, 30 vs. 70, or 50 vs. 50) for each choice. During the BART, participants were told that they would inflate a balloon on each of a number of trials by pressing “pump” button. Each pump could earn participants 0.1 CNY; however, if the balloon exploded, they would lose the money accumulated on that trial. Older adults reported lower levels of impulsive behavior in motor impulsiveness, attentional impulsiveness, and non-planning impulsiveness as compared with younger adults. However, older adults were more likely to choose options that had a low probability of winning on the CGT, showing as a risk-seeking propensity. The decision of older adults were also of significantly differences than those of younger adults on the BART; however, for this task, older adults made more risk averse decisions than the younger adults. Besides the heterogeneous results of these two tasks, the findings had showed that older adults made significantly lower quality decisions than younger adults. Moreover, the moderation analysis showed that the effects of age upon decision making reflect an indirect influence of age-related impulsiveness decline. In conclusion, different decision making tasks do not entirely succeed in capture the same age-related risk-taking propensity. Moreover, our study highlights that the relationship between aging and decision making is mediated, in part, by self-assessment of impulsiveness. Our results show that in order to truly understand the age-related alterations upon decision making, future research may consider take into account the convergent and predictive validity of self-report and behavioral measures of risk taking.