It was the objective of this study to investigate the relation between vitronectin and plasminogen activator inhibitor (PAI)-1 plasma levels with nine-year incidences of the metabolic syndrome (MetS) and of type 2 diabetes mellitus (T2DM). Baseline plasma concentrations of vitronectin and PAI-1 were measured in 627 healthy participants from the prospective D.E.S.I.R. cohort who subsequently developed MetS (n = 487) and T2DM (n = 182) over a nine-year follow-up (42 presented both) and who were matched with two healthy control subjects each by use of a nested case-control design. Parameters composing the MetS explained about 20% of plasma vitronectin levels. An increase of one standard deviation of vitronectin was associated with increased risks of both the MetS (odds ratio [OR] = 1.21 [1.07 - 1.37], p = 0.003) and T2DM (OR = 1.24 [1.01 - 1.53], p = 0.045). Corresponding ORs for PAI-1 levels were 1.46 [1.27 - 1.68] (p<10(-4)) and 1.40 [1.14 - 1.72] (p = 0.0012). However, the effects of vitronectin and PAI-1 levels on outcomes were not independent. The vitronectin-MetS association was restricted to individuals with low to modest PAI-1 levels (OR = 1.33 [1.14 - 1.54], p = 0.0003) while no association was observed in individuals with high PAI-1 levels (OR = 0.87 [0.68 - 1.10], p = 0.24), the test for interaction being highly significant (p = 0.0009). In conclusion, baseline plasma vitronectin is a marker of incident MetS at nine years. Its predictive ability for MetS and T2DM should not be assessed independently of PAI-1 levels.