Device-guided, home-based slow deep breathing (SDB) training has been shown to lower blood pressure (BP) in some patients with primary hypertension, however, the underlying mechanisms remain obscure. We sought to determine the influence of SDB training on BP, central sympathetic outflow and arterial baroreflex function in patients with primary hypertension. In addition, the effect of SDB training on arterial stiffness, cardiac structure and function were assessed also. Nineteen patients with essential hypertension (mean±SD: age 55 ± 12 years, 11 males, body mass index 28 ± 2.8 kg/m 2 ; median [interquartile range]: duration of hypertension 48 [12–169] month) were recruited. Medication use (58% calcium channel blockers, 42% angiotensin-converting-enzyme inhibitor or angiotensin receptor blockers, 37% thiazide diuretics, 21% beta-blockers and 26% statins) was stable before (3 months) and during the study. Patients were studied before and after 8-weeks of home-based SDB training (10 min/day, 4 times/week) with a RESPERATE O device. On study days, BP (automated sphygmomanometry), respiration (oronasal mask, heated pneumotach), heart rate (HR; electrocardiogram) and muscle sympathetic nerve activity (SNA; microneurography) were measured for 10 min while patients rested in a supine position. Indices of spontaneous cardiac baroreflex sensitivity (sequence technique) and arterial baroreflex control of the muscle SNA were calculated. A measure of central arterial stiffness (aortic augmentation index; AI) was determined using a Sphygmocor device. Cardiac structure and function were assessed by echocardiography. The SDB training decreased office systolic BP from 147 ± 18 to 133 ± 17 mmHg (mean±SD; P = 0.001), diastolic BP from 84 ± 10 to 79 ± 12 mmHg (P = 0.008) and pulse pressure from 63 ± 15 to 55 ± 13 mmHg (P = 0.003). In addition, SDB training significantly reduced muscle SNA burst frequency from 33.8 ± 8.2 to 28.3 ± 7.1 bursts/min (P = 0.03) and burst incidence (i.e., percentage of cardiac cycles associated with a sympathetic burst) from 59.0 ± 19.5 to 48.6 ± 14.4 bursts/100 heartbeats (P = 0.05). Respiratory frequency, tidal volume, and HR were unchanged (61 ± 9 vs. 61 ± 7 beats/min, P = 0.83). Neither spontaneous cardiac baroreflex sensitivity (systolic BP–RR interval index, 9.4 [6.93–14.83] to 9.5 [5.8–14.2] ms/mmHg, P = 0.71) nor arterial baroreflex control of muscle SNA (-2.9 [-5.1 – -1.3] to -2.4 [-7.0 – -1.1] bursts/100 heartbeats/mmHg, P = 0.58) were changed following SDB training. No significant changes in AI, cardiac structure and function were detected following SDB training. Our findings indicate for the first time that reductions in BP following 8-weeks of device-guided, home-based SDB training are accompanied by reductions in central sympathetic outflow in patients with primary hypertension, while arterial baroreflex sensitivity, central arterial stiffness, cardiac structure and function remain unchanged. This work is funded by British Heart Foundation (PG/11/41/28893).