Alterations in the secretion of adipokines may explain the link between obesity, type 2 diabetes (T2DM) and coronary artery disease (CAD). These conditions have been associated with variation in the adiponectin gene, although evidence for this relationship has been variable, with differences found even in similar samples. This study aims to clarify these inconsistencies by determining the impact of identified adiponectin gene (ADIPOQ) variants (-11391G>A,-1377C>G[promoter] and +45T>G[exon 2] and +276G>T[intron 2]) on the prospective risk of CAD and T2DM in healthy men, and on adverse metabolic markers, in myocardial infarct survivors and controls from different parts of Europe. The hazard ratio for cardiovascular disease varied across the -11391GG/GA/AA(p = 0.03) and -11371CC/CG/GG(p = 0.05) genotypes only. In contrast, only the +45T>G variant (3.80[1.76-8.24]) was associated with T2DM, while two haplotypes GCTT/GCGG (p < 0.05) and +276G>T(p = 0.01) increased risk in interaction with obesity. The variants were associated with a number of biomarkers in Southern but not Northern Europe (p = 0.01), despite no significant differences in allele or haplotype frequencies (p > 0.44). A risk haplotype could not be identified in either sample. Adiponectin gene variants are hence currently poor markers for the development of T2DM and CAD. Their influence on risk depends significantly on interactions that are not currently understood with either genetic variation elsewhere or the environment of the sample studied.