Introduction Depressed older adults are at high-risk for suicidal behavior (Pfeiffer et al., 2013, Vannoy et al., 2007) and for developing dementia (Kaup et al, 2016; Mizra et al., 2016). Executive dysfunction has been described across the life-span in suicide attempters (Keilp et al., 2013, Clark et al., 2011, Richards-Devantoy et al., 2012). However, few studies have explored the links between several domains of cognition and suicide risk in those with late-life depression and the extant evidence is equivocal at best (Conejero et al., 2018, Tu et al., 2018, Erlangsen et al., 2008). Suicidal behavior is heterogeneous, and it is unclear whether the medical seriousness (Beck et al., 1975) of the suicide attempt or whether the onset of first suicidal behavior (initiated earlier in life relative to late-life (60+ years)) relates differently to cognitive deficits. Methods This project uses a case-control design and includes 278 adults aged 50?years and older: 56 non-depressed controls (no life-time history of psychiatric disorder), 67 non-suicidal depressed older adults(no life-time history of suicide ideation or suicide attempt), 63 depressed suicide ideators, and 92 depressed adults with a history of suicide attempt. Analysis 1 examined cognitive profiles of low (n=52) vs. high-lethality (i.e., potentially fatal without medical intervention) (n=40) suicide attempters relative to non-suicidal depressed older adults and non-depressed controls. Analysis 2 examined cognitive profiles of those making their first suicide attempt prior to age 60 (early onset: n=66) and those making their first suicide attempt at or after age 60 (late-onset: n=26) relative to non-suicidal depressed older adults and non-depressed controls. Lethality of attempt was similarly distributed across early- and late-onset attempters. Depression severity was assessed using the Hamilton Depression Rating Scale (HAM-D-17 minus the suicide item). Cognitive functioning was assessed using the Repeatable Battery of Neuropsychological Status (RBANS) and the Trail Making Test from the Delis Kaplan Executive Functioning System (D-KEFS). Linear mixed effects models were used for assessing group differences cognitive domains assessed using multiple subtests or trials (i.e., learning and memory) and linear regression was used for cognitive domains assessed using a single test (i.e., processing speed, executive functioning). Demographic covariates were used for all data analyses (i.e., age, sex, education). Analyses comparing suicidal subgroups with non-suicidal depressed older adults additionally adjusted for depression severity and lifetime history of anxiety and substance use disorders. Results High-lethality attempters performed worse than low-lethality attempters and non-depressed controls on measures of verbal memory (Linear mixed effects model; group x test interaction: F=1.997, p=0.03). On a measure of executive functioning, low- but not high-lethality attempters performed worse than non-suicidal depressed older adults (LL: Beta= -0.18, p=0.01; HL: Beta=-0.12, p=0.08) and non-depressed controls (LL: Beta=0.31, p Conclusions Cognitive profiles of older suicidal attempters varied based on the lethality of most lethal suicide attempt and age of first suicide attempt. Those with low-lethality and early-onset suicide attempts exhibited executive dysfunction relative to non-suicidal depressed older adults, in-line with previous findings indicating executive dysfunction as part of life-long suicide vulnerability factors. In contrast, late-onset attempters exhibited slower processing speed relative to non-suicidal depressed older adults, and high-lethality attempters performed worse on memory measures relative to low-lethality attempters, possibly indicating a dementia prodrome. This research was funded by: NIMH: R01 MH085651; VA VISN 4 MIRECC Advanced Research Fellowship, VA Office of Academic Affiliations