The voltage sensor dynamics that regulate voltage-gated ion channels are of significant interest to the understanding of bioelectric signals. Recent studies have shown that prolonged activation of voltage-gated potassium (Kv) channels leads to a reconfiguration of the voltage sensor, termed relaxation (Lacroix et al., 2011). Relaxation imparts hysteresis to the voltage sensor, such that the voltage-dependence of activation and deactivation differ. Voltage sensor hysteresis was recently demonstrated in hERG channels (Tan et al., 2013; Hull et al., 2014), where relaxation of the voltage sensor may produce physiologically relevant slowing of channel deactivation. Slow deactivation of hERG channels, along with slow activation and rapid onset and recovery from inactivation, is critical for their role in cardiac repolarization. This is highlighted by the observation that mutations that cause accelerated deactivation are associated with Long QT Syndrome and sudden cardiac death (Chen et al., 1996). The mechanism underlying voltage sensor relaxation in hERG channels was investigated using two-electrode voltage clamp, cut-open vaseline gap clamp and voltage clamp fluorimetry in hERG channels expressed in X. laevis oocytes (terminal anaesthesia; 2 g/l tricaine) to document ionic current and voltage sensor movement during channel activation and deactivation. Increasing the duration of the activating voltage step was shown to slow the kinetics of hERG channel closure. The effect of activation pre-pulse duration on the rate of deactivation at -110 mV could be described by a single exponential with a tau value of 216 ± 31 ms (n = 5). Allowing relaxation to approach steady-state (i.e. following a 500 ms activating pre-pulse) resulted in a shift of the voltage dependence of deactivation -30 mV along the voltage axis with respect to that of activation (V1/2act = -27.1 ± 1.2 mV vs V1/2deact = -57.5 ± 4.2 mV; n = 5). A chimeric approach was used to assess the role of S3-S4 linker length, since it has been reported to mediate the extent of voltage sensor relaxation in Shaker channels (Priest et al., 2013). Chimeric hERG channels in which the native 9 residue S3-S4 linker was replaced with the 31 residue linker of Shaker showed wild-type-like hysteresis, suggesting that S3-S4 linker length is not an important mediator of relaxation in hERG as proposed for Shaker. An alternative approach showed that acidic external pH reduced voltage sensor relaxation with a pKa of 5.5 (n = 5). Low pH (pH 5.5) reduced the separation of the voltage dependence of activation and deactivation as well as the fluorescence report of voltage sensor hysteresis, which was reduced by 67 % to -16 ± 2.5 mV (n = 5). These data suggest that voltage sensor relaxation is directly modified by external protons and that acidic pH destabilizes the relaxed conformation of the voltage sensor resulting in accelerated deactivation kinetics.