OBJECTIVE: To investigate the utility and accuracy of radial-aortic arterial transfer functions for the derivation of central blood pressure waveforms. DESIGN Prospective measurement of central and peripheral waveforms in patients undergoing coronary angiography or percutaneous coronary intervention. METHODS: Simultaneous invasive central aortic and non-invasive radial pressure waveforms were recorded in 78 subjects (61 male : 17 female). Data were applied to a single-input/single-output model for the calculation of a transfer function (TF). Individual TFs were derived by two methods and ensemble averaged TFs obtained for the group. Reverse transformation was performed using each averaged TF applied to the radial data of each subject. RESULTS: There was close linear correlation between measured aortic parameters and both radial and TF-derived aortic systolic and diastolic pressures (P < 0.001) and most other waveform parameters. However, despite small mean differences between measured and most TF-derived aortic parameters (systolic pressure 0.8-2.9 mmHg, augmentation index 4.3-5.6%), individual scatter was marked, with 95% limits of agreement of +/- 14.6 mmHg (systolic pressure) and +/- 24.4% [augmentation index (AI)]. Indeed, scatter for AI was so marked that measured and derived values were not statistically significantly correlated. CONCLUSIONS: Transfer functions may be valid for the derivation of some central aortic waveform characteristics. However, in providing neither improved reproducibility nor data on parameters not obtainable from the radial waveform, transfer function techniques may offer no additional clinical benefit. The absence of correlation between measured and TF-derived aortic AI and wide limits of agreement of other parameters should be considered if this technique is utilized in clinical practice.