Background Advanced glycation endproducts (AGEs) have been linked to cardiovascular disease (CVD) in cohorts with and without diabetes. Data are lacking on prospective associations of various α‐dicarbonyl‐derived AGEs and incident CVD in the general population. We tested the hypothesis that major plasma AGEs are associated with new‐onset CVD in 2 population‐based cohorts of differing age and comorbidities. Methods and Results Analyses involved a random subcohort (n=466) from the Cardiovascular Health Study and a case‐cohort sample (n=1631) from the Multi‐Ethnic Study of Atherosclerosis. Five AGEs and 2 oxidative products were measured by liquid chromatography tandem mass spectrometry. Associations with CVD (myocardial infarction and stroke) were evaluated with Cox regression. Participants in the Cardiovascular Health Study were older than the Multi‐Ethnic Study of Atherosclerosis, and had more comorbidities, along with higher levels of all AGEs. During median follow‐up of 11 years, 439 participants in the Multi‐Ethnic Study of Atherosclerosis and 200 in the Cardiovascular Health Study developed CVD. After multivariable adjustment, carboxymethyl‐lysine, 3‐deoxyglucosone hydroimidazolones and a summary variable of all measured AGEs (principal component 1) were significantly associated with incident CVD in the Cardiovascular Health Study (HRs [95% CI]: 1.20 [1.01, 1.42], 1.45 [1.23, 1.72], and 1.29 [1.06, 1.56], respectively), but not the Multi‐Ethnic Study of Atherosclerosis. Oxidative products were not associated with CVD in either cohort. Conclusions We found α‐dicarbonyl‐derived AGEs to be associated with CVD in an older cohort, but not in a healthier middle‐aged/older cohort. Our results suggest that AGEs may exert detrimental cardiovascular effects only under conditions of marked dicarbonyl and oxidative stress. Further investigation of α‐dicarbonyl derivatives could lead to potential new strategies for CVD prevention in high‐risk older populations.