Abstract Aims/hypothesis Reactive α-dicarbonyl compounds are major precursors of AGEs and may lead to glycation of circulating and/or cell-associated complement regulators. Glycation of complement regulatory proteins can influence their capacity to inhibit complement activation. We investigated, in a human cohort, whether greater dicarbonyl stress was associated with more complement activation. Methods Circulating concentrations of dicarbonyl stress markers, i.e. α-dicarbonyls (methylglyoxal [MGO], glyoxal [GO] and 3-deoxyglucosone [3-DG]), and free AGEs (N^ε-(carboxymethyl)lysine [CML], N^ε-(carboxyethyl)lysine [CEL] and N^δ-(5-hydro-5-methyl-4-imidazolon-2-yl)-ornithine [MG-H1]), and protein-bound AGEs (CML, CEL, pentosidine), as well as the complement activation products C3a and soluble C5b-9 (sC5b-9), were measured in 530 participants (59.5 ± 7.0 years [mean ± SD], 61% men) of the Cohort on Diabetes and Atherosclerosis Maastricht (CODAM) study. Multiple linear regression analyses were used to investigate the associations between dicarbonyl stress (standardised) and complement activation (standardised) with adjustment of potential confounders, including age, sex, lifestyle, use of medication and markers of obesity. In addition, the associations of two potentially functional polymorphisms (rs1049346, rs2736654) in the gene encoding glyoxalase 1 (GLO1), the rate-limiting detoxifying enzyme for MGO, with C3a and sC5b-9 (all standardized) were evaluated. Results After adjustment for potential confounders, plasma concentration of the dicarbonyl GO was inversely associated with sC5b-9 (β −0.12 [95% CI –0.21, −0.02]) and the protein-bound AGE CEL was inversely associated with C3a (−0.17 [−0.25, −0.08]). In contrast, the protein-bound AGE pentosidine was positively associated with sC5b-9 (0.15 [0.05, 0.24]). No associations were observed for other α-dicarbonyls and other free or protein-bound AGEs with C3a or sC5b-9. Individuals with the AG and AA genotype of rs1049346 had, on average, 0.32 and 0.40 SD lower plasma concentrations of sC5b-9 than those with the GG genotype, while concentrations of C3a did not differ significantly between rs1049346 genotypes. GLO1 rs2736654 was not associated with either C3a or sC5b-9. Conclusions/interpretation Plasma concentrations of dicarbonyl stress markers showed distinct associations with complement activation products: some of them were inversely associated with either C3a or sC5b-9, while protein-bound pentosidine was consistently and positively associated with sC5b-9. This suggests different biological relationships of individual dicarbonyl stress markers with complement activation.