Background: Long QT Syndrome type 3 (LQT3) is caused by mutations in SCN5a, which encodes the hNa V 1.5 α subunit; each contains 4 domains (DI-DIV) with 6 transmembrane segments (S1-S6). S1-S4 form a voltage sensing domain (VSD) that transduces membrane potential into channel gating. The molecular mechanisms that lead to LQT3 and the consequences of mutation location remain unclear, and better definition will lead to improved phenotypic precision and therapeutic targeting. We use a recent technique, voltage-clamp fluorometry (VCF), to evaluate three LQT3 mutations in the DIV VSD: R1623Q, R1626P, and previously uncharacterized, S1609W. Methods: To track the S4, a cysteine was introduced into the DIV S3-S4 linker, and conjugated to a TAMRA-MTS label. mRNA was injected into Xenopus oocytes and currents and fluorescence were recorded after 3-5 days with cut-open VCF. Results: Activation, inactivation, and recovery were similar in WT and S1609W. R1626P activation and inactivation recovery was not significantly different from WT while inactivation was left shifted. Both mutants exhibited a pro-arrhythmic delay of inactivation onset at -30mV - S1609W by 44% (0.8 ms) and R1626P by 36% (0.7 ms), when compared to onset in WT (1.9 ± 0.0 ms). DIV S4 deactivation also differed significantly among WT, S1609W, and R1626P. Exponential fitting revealed two time constants for the WT DIV S4 to return to rest of 0.23 ms and 17.44 ms with magnitudes of 0.3 and 0.7 ± 0.1, respectively. In contrast, S1609W showed a significant increase in the fast component (0.8 and 0.2 ± 0.1). A single signal from R1626P showed an intermediate phenotype (0.7 and 0.3). Further, a delay in sensor deactivation onset observed in the WT was absent from both mutants. No fluorescence was detected from R1623Q, perhaps indicating an especially severe phenotype with respect to the DIV S4 motion. Conclusion: Fluorescent kinetics from two LQT3 mutants, S1609W and R1626P showed a progressive molecular phenotype in which a slow component of recovery to the resting state was impaired by the mutation. Comparison to ionic current inactivation revealed that the reduction in this slow component corresponded to an increase in the time to peak, implying that the onset of fast inactivation is delayed by these mutations.