Background:Different methods to measure carotid–femoral pulse wave velocity (CFPWV) may affect the measurements obtained and influence the association between CFPWV, cardiovascular risk factors and biomarkers of subclinical vascular health. The estimation of distance between the carotid and femoral artery measurement sites (the arterial path length) is particularly problematic.Method:We determined if CFPWV and equation-based estimates of CFPWV were influenced by arterial path length and if this affected the association of CFPWV with cardiovascular risk factors and subclinical vascular biomarkers. The CFPWV derived from the measurement of surface distance (CFPWV-D), arterial path length formula (CFPWV-F), and estimated CFPWV (ePWV) were obtained from 489 older adults (67.2 ± 8.8 years). Macrovascular [carotid artery: lumen diameter (LD), inter-adventitial diameter (IAD), intima–media thickness (IMT) and total plaque area (TPA)] and microvascular [reactive hyperaemia index and urinary albumin-creatinine ratio (UACR)] biomarkers were also measured.Results:CFPWV-D was significantly greater than CFPWV-F [9.6 (8.0–11.2) vs. 8.9 (7.6–10.5) m/s, P < 0.001], because of estimated path length being longer in CFPWV-D than CFPWV-F (495.4 ± 44.8 vs. 465.3 ± 20.6 mm, P < 0.001). ePWV was significantly greater than both CFPWV-F and CFPWV-D [11.0 (10.0–12.2) m/s, P < 0.001]. The three CFPWV methods were similarly associated with LD, IAD, IMT, TPA and UACR but not with cardiovascular risk factors.Conclusion:Different methods to measure CFPWV affect the derived measurement values and the association with cardiovascular risk factors but not the association with subclinical biomarkers of vascular health. These hitherto unreported observations are important considerations in experimental design, data interpretation and of particular importance, comparison between studies where CFPWV is measured.