The present study assessed whether increased fruit and vegetable (F&V) intake reduced the concentrations of the inflammatory marker serum amyloid A (SAA) in serum, HDL2 and HDL3 and whether the latter reduction influenced any of the functional properties of these HDL subfractions. The present study utilised samples from two previous studies: (1) the FAVRIT (Fruit and Vegetable Randomised Intervention Trial) study – hypertensive subjects (systolic blood pressure (BP) range 140–190 mmHg; diastolic BP range 90–110 mmHg) were randomised to receive a 1-, 3- or 6-portion F&V/d intervention for 8 weeks, and (2) the ADIT (Ageing and Dietary Intervention Trial) study – older subjects (65–85 years) were randomised to receive a 2- or 5-portion F&V/d intervention for 16 weeks. HDL2 and HDL3 were isolated by rapid ultracentrifugation. Measurements included the following: serum high-sensitive C-reactive protein (hsCRP) by an immunoturbidimetric assay; serum IL-6 and E-selectin and serum-, HDL2- and HDL3-SAA by ELISA procedures; serum-, HDL2- and HDL3-cholesterol ester transfer protein (CETP) activity by a fluorometric assay. Although the concentrations of hsCRP, IL-6 and E-selectin were unaffected by increasing F&V intake in both studies (P>0·05 for all comparisons), those of SAA in HDL3 decreased in the FAVRIT cohort (P= 0·049) and those in HDL2 and HDL3 decreased in the ADIT cohort (P= 0·035 and 0·032), which was accompanied by a decrease in the activity of CETP in HDL3 in the FAVRIT cohort (P= 0·010) and in HDL2 in the ADIT cohort (P= 0·030). These results indicate that SAA responds to increased F&V intake, while other inflammatory markers remain unresponsive, and this leads to changes in HDL2 and HDL3, which may influence their antiatherogenic potential. Overall, the present study provides tangible evidence of the effectiveness of increased F&V intake, which may be of use to health policy makers and the general public.