Beat-to-beat variability in blood pressure (BP) is associated with recurrent stroke despite good control of hypertension. However, no study has identified rates of progression of beat-to-beat BP variability (BPV), its determinants, or which patient groups are particularly affected, limiting understanding of its potential as a treatment target. In consecutive patients one month after a transient ischaemic attack or nondisabling stroke (Oxford Vascular Study), continuous noninvasive BP was measured beat-to-beat over 5 minutes (Finometer). Arterial stiffness was measured by carotid-femoral pulse wave velocity (Sphygmocor). Repeat assessments were performed at the 5-year follow-up visit and agreement determined by intraclass correlation coefficient. Rates of progression of systolic BPV (SBPV) and diastolic BPV (DBPV) and their determinants were estimated by mixed-effect linear models, adjusted for age, sex, and cardiovascular risk factors. One hundred eighty-eight of 310 surviving, eligible patients had repeat assessments after a median of 5.8 years. Pulse wave velocity was highly reproducible but SBPV and DBPV were not (intraclass correlation coefficient: 0.71, 0.10, and 0.16, respectively), however, all 3 progressed significantly (pulse wave velocity, 2.39%,
P
<0.0001; SBPV, 8.36%,
P
<0.0001; DBPV, 9.7,
P
<0.0001). Rate of progression of pulse wave velocity, SBPV, and DBPV all increased significantly with age (
P
<0.0001), with an increasingly positive skew and were particularly associated with female sex (pulse wave velocity
P
=0.00035; SBPV
P
<0.0001; DBPV
P
<0.0001) and aortic mean SBP (SBPV
P
=0.037, DBPV
P
<0.0001). Beat-to-beat BP variability progresses significantly in high-risk patients, particularly in older individuals with elevated aortic systolic pressure. Beat-to-beat BPV and its progression represent potential new therapeutic targets to reduce cardiovascular risk.