The objective of this study was to compare central systolic blood pressure (cSBP) and augmentation index (AIx) from two recently introduced devices, Omron HEM-9000 (OM) and Arteriograph (AG), not using a transfer function with those of the widely used SphygmoCor (SC) calibrated on brachial blood pressure like OM. Random-order manufacturer-recommended measurements using SC and OM by radial tonometry and AG were taken on the left arm in 35 men (54±10 years) after 5 min supine rest. Results are means (95% confidence interval) of differences using paired t-tests. cSBP by OM was 4.1 (1.0-7.1) mm Hg higher than by AG. Both OM and AG estimated the mean cSBP to be significantly higher than did SC (114.8 mm Hg) by 12.5 (10.3-14.7) and 8.6 (4.9-12.3) mm Hg, respectively, although closely correlating with SC (r¼0.9). Calibrating SC with diastolic blood pressure (DBP) and more accurate mean arterial pressure (as DBP+0.4ÂPP) resulted in significantly higher cSBP statistically not different from AG's cSBP: 0.9 (À1.1 to +2.9)mm Hg, and closer to OM's: 5.1 (3.4-6.8)mm Hg. Radial AIx from SC and OM disagreed by 3 (0.7-5.4)%, and correlated (r¼0.8) with AG's brachial AIx. AG's aortic AIx was 7.9 (5.7-10.2)% higher than SC's, but closely correlated (r¼0.9). Clinically significant, higher cSBP measured by AG, OM and more accurately calibrated SC adds to previous data suggesting that SC measurements by classic calibration underestimate cSBP. Invasive studies involving all three devices would be more illuminating.