<h3>Introduction</h3> There is increasing clinical and legal interest in the decision-making abilities of older adults given the potentially far-reaching consequences, particularly with respect to financial well-being. For example, the Federal Bureau of Investigation's most recent elder fraud report showed that 105,301 individuals over the age of 60 fell victim to fraud in 2020, resulting in total losses of nearly $1 billion. This highlights the importance of investigating potential factors associated with changes in risk/reward financial decision-making in older adults. The reasons behind age-related changes in financial decision-making are still poorly understood, though several factors have been proposed, such as age-related affective changes (Kensinger & Leclerc, 2009). In addition, late-life mood disorders, such as major depression and bipolar disorder, may further alter or compromise financial decision-making (Fein et al., 2007). These findings have been inconsistent, however, and this remains an understudied area. Accordingly, the aims of this study were twofold: (1) To investigate risk/reward financial decision-making in older adults with and without mood disorders using a well-validated paradigm, the Iowa Gambling Task (IGT), and (2) To determine whether the relationship between group and IGT performance varied as a function of positive or negative affective state, as measured by the Positive and Negative Affect Scale (PANAS). <h3>Methods</h3> Forty-five older adults ranging in age from 55 to 86 (<i>M</i>=68.07 ± 8.14) were selected from the Geriatric Mood Disorders Database (GMDD) Study at McLean Hospital (19 healthy control, 26 mood disorder). The mood disorder group included patients with bipolar disorder (<i>n</i>=13) or major depressive disorder (<i>n</i>=14). For aim 1, a Group (control vs mood) by Trial Block (1-5) repeated measures ANOVA was performed using IGT net score as the outcome variable. Independent samples t-tests were then used to assess group differences in overall deck preference (i.e., risky vs conservative decks) and sensitivity to punishment frequency. For aim 2, linear regression was performed using Group, PANAS scores, and Group x PANAS interaction terms as predictors of IGT performance. <h3>Results</h3> Groups did not differ in age, education (∼16 years in both groups), or global cognitive ability (MMSE). The mood group endorsed more negative affectivity on the PANAS [<i>p</i> = .001], whereas groups did not differ in positive affectivity. Repeated measures ANOVA showed a marginally significant main effect of Trial Block [<i>F</i>(1,4) = 2.40, <i>p</i> = 0.07], reflecting a general trend toward more conservative or advantageous choices over time. However, this trend did not vary by Group [<i>p</i> = 0.22]. The groups also did not differ in deck preference [all <i>p</i>s > .19] or sensitivity to punishment frequency [<i>p</i> = 0.58]. Regression analysis revealed a marginally significant Group x PANAS-Neg interaction in Trial Block 1 Net Score [<i>β</i> = -3.90, <i>t</i>(43) = -1.95, <i>p</i> = 0.059], which reflected a significant positive correlation between negative affectivity and more conservative or advantageous decision making on early trials in healthy controls only [Spearman's ρ = 0.56, <i>p</i> = 0.13]. A marginally significant Group x PANAS-Pos interaction was also found for Deck A preference – i.e., one of the risky decks [<i>β</i> = 0.54, <i>t</i>(43) = 1.89, <i>p</i> = 0.067]. This reflected a positive correlation between Deck A choices and positive affectivity in the mood group, but the opposite in the control group, though neither correlation was significant [<i>p</i>s > .10]. There were no Group x PANAS interactions in sensitivity to punishment frequency. <h3>Conclusions</h3> This study showed no differences between healthy older adults and those with mood disorders in IGT performance, with both groups exhibiting a trend toward more advantageous (i.e., less risky) choices over time. This suggests the presence of a mood disorder in later life may not compromise one's ability to integrate feedback and improve risk/reward decision-making over time in a financial context. However, this conclusion should be interpreted cautiously given the somewhat modest sample size and generally high education level in both groups, with the latter potentially being protective in some way. With those same caveats in mind, interesting trend-level group interactions were also seen when the affective state was considered, suggesting a potential differential impact of positive and negative affectivity on decision-making style across groups. The possible effects of affective valence on risky decision-making in older adults should be investigated in larger and more diverse samples to help clarify the reliability and generalizability of the current findings.