Recent trials support the efficacy of renal sympathetic denervation (RDN) to reduce blood pressure (BP). Nevertheless, about one third of patients are considered non-responders to RDN. Previous retrospective analyses suggest arterial stiffness could predict BP response to RDN. We prospectively assessed the potential of invasive pulse wave velocity (iPWV) to predict BP response to RDN. Additionally, we aimed to establish non-invasive models based on arterial stiffness to predict BP response to RDN. iPWV, magnetic resonance imaging-based markers of arterial stiffness and the carotid-femoral pulse wave velocity were recorded prior to RDN in patients with treatment resistant hypertension. Changes in daytime BP after 3 months were analysed according to the prespecified iPWV cut-off (14.4 m/s). Regression analyses were used to establish models for non-invasive prediction of BP response. Results were compared to iPWV as reference and were then validated in an external patient cohort. Eighty patients underwent stiffness assessment before RDN. After 3 months, systolic 24h and daytime BP were reduced by 13.6±9.8 mmHg and 14.7±10.6 mmHg in patients with low iPWV, versus 6.2±13.3 mmHg and 6.3±12.8 mmHg in those with high iPWV (p<0.001 for both). Upon regression analysis, logarithmic ascending aortic distensibility and systolic baseline BP independently predicted BP change at follow-up. Both were confirmed in the validation cohort. iPWV is an independent predictor for BP response after RDN. In addition, BP change prediction following RDN using non-invasive measures is feasible. This could facilitate patient selection for RDN treatment.