Voltage-gated sodium channel (VGSC) blockers have clinical applications for chronic pain states and parasthesiae that arise from aberrant action potential firing in sensory neurons (Lai et al. 2003). Flecainide (FLEC), a VGSC blocker, is an anti-arrhythmic drug that has been found to have beneficial effects in animal models of neuropathic pain and inflammatory demyelinating disease (Ichimata et al. 2001; Bechtold et al. 2004). Here, we describe effects of FLEC on compound action potential (CAP) conduction in sensory nerve that may contribute to these actions. Rat saphenous nerves (n = 25) were isolated following humane killing and CAPs were recorded in vitro using a grease-gap method. FLEC (10 μM-1 mM) reduced CAP amplitudes dose-dependently with IC₅₀ values of 484.6 ± 78.0 μM (mean ± SEM, , n = 4 for 10 μM, n = 6 for 30 μM – 1 mM) in A fibres and 109.7 ± 27.2 μM (n = 7) in C fibres. Increasing extracellular potassium from 3 mM to either 8 mM (A fibres) or 5 mM (C fibres) reduced the IC₅₀ values to 138.5 ± 19.5 μM (n = 5 for 10 and 300 μM, n = 7 for 30 and 100 μM) for A fibres and 67.1 ± 17.4 μM (n = 5) for C fibres. We have previously observed similar results with the anticonvulsants carbamazepine (CBZ) and lamotrigine (LTG) in the same preparation; these results are consistent with a voltage-dependent block of VGSCs. During impulse trains (100 CAPs for A fibre and 50 for C fibre) FLEC showed a frequency-dependent block of CAPs that was not seen previously with CBZ or LTG. At 100 μM FLEC in A fibres, the inhibition of CAP amplitude between the first and last CAP in the train increased by 5 ± 1 % (paired t-test p = 0.003, n = 6) and 19 ± 2 % (p = 0.002, n = 6) at 1 and 10 Hz respectively in 3 mM potassium buffer and 5 ± 1 % (p = 0.0003, n = 7) and 16 ± 1 % (p < 0.0001, n = 7) when potassium was increased to 8 mM. With C fibres at 100 μM FLEC, increased inhibition at higher frequency was seen in 3 mM (1 Hz: 31 ± 5 %, p = 0.001, n = 6; 10 Hz: 39 ± 7 %, p = 0.005, n = 6) but not in 5 mM potassium buffer. Surprisingly, despite this frequency-dependent blocking effect FLEC, like CBZ and LTG, had no effect of the refractoriness of A fibres measured using a paired pulse protocol. The frequency-dependent blocking effect that distinguishes FLEC from CBZ and LTG may be due to different mechanisms of VGSC block. CBZ and LTG bind predominantly to an inactivated state of the VGSC (Rogawski & Löscher 2004) whereas FLEC probably acts primarily via an open channel block (Ramos & O′Leary, 2004). If there is insufficient time for drug dissociation from channels between impulses then VGSCs blocked by FLEC may accumulate thereby increasing the inhibitory effect of the drug.