The purpose of this study was to measure changes in plasma adiponectin (ApN) over 24 months of exercise intervention in middle age adults with a predisposition to metabolic syndrome and to determine if changes in ApN were more affected by physical activity or physical fitness. Thirty-six subjects completed a 24 months home-based exercise program (cycling>or=three times per week, >or=45 min/session at 50-65% of VO2peak). Body composition, blood samples, and physical fitness were studied at baseline and after 12 and 24 months of participation in the study. The prescribed physical activity was monitored via self-reported exercise diary to determine MET levels, hours, and exercise compliance. Two-tailed repeated measures ANOVA and Spearman Rank Correlation Coefficients were used to detect significant differences and associations between the variables. ApN increased significantly (P<0.05) after 12 months in males (n=17; 5.3+/-1.9-7.0+/-3.0 microg ml-1) but not in females (n=9; 8.6+/-3.8-11.5+/-4.0 microg ml-1). The net change in ApN over 24 months was significantly correlated to the net change in VO2peak (physical fitness) (r=0.66; P<0.001), whereas exercise intensity was negatively correlated to DeltaApN over 12 months (r=-0.4; P<or=0.04) and 24 months (r=-0.45; P<or=0.02). Based on our results, an improvement in cardiorespiratory fitness of 15% increased plasma ApN concentration. Our findings suggest that moderate physical activity performed over many months induces positive changes in the plasma ApN concentration in adults with a predisposition to metabolic syndrome.