Abstract: Sleep problems are common among veterans with post-traumatic stress disorder and closely associated with hyperarousal symptoms. Transcutaneous vagus nerve stimulation (tVNS) may have potential to improve sleep quality in veterans with PTSD through effects on brain systems relevant to hyperarousal and sleep-wake regulation. The current pilot study examines the effect of 1 h of tVNS administered at "lights out" on sleep architecture, microstructure, and autonomic activity. Thirteen veterans with PTSD completed two nights of laboratory-based polysomnography during which they received 1 h of either active tVNS (tragus) or sham stimulation (earlobe) at "lights out" with randomised order. Sleep staging and stability metrics were derived from polysomnography data. Autonomic activity during sleep was assessed using the Porges-Bohrer method for calculating respiratory sinus arrhythmia (RSA(P-B) ). Paired t-tests revealed a small decrease in the total sleep time (d = -0.31), increase in N3 sleep (d = 0.23), and a small-to-moderate decrease in REM sleep (d = -0.48) on nights of active tVNS relative to sham stimulation. tVNS was also associated with a moderate reduction in cyclic alternating pattern (CAP) rate (d = -0.65) and small-to-moderate increase in RSA(P-B) during NREM sleep. Greater NREM RSA(P-B) was associated with a reduced CAP rate and NREM alpha power. This pilot study provides preliminary evidence that tVNS may improve sleep depth and stability in veterans with PTSD, as well as increase parasympathetically mediated nocturnal autonomic activity. These results warrant continued investigation into tVNS as a potential tool for treating sleep disturbance in veterans with PTSD.