Rational drug design for modulators of membrane-embedded proteins is traditionally guided by the lock-and-key approach, i.e. the transmembrane protein comprises an active site (lock), to which only a ligand (key) with specific structural features can fit to influence activity of the protein. However, the lipid membrane surrounding target proteins is often overlooked in the lock-and-key mechanism. This work seeks to understand the role that the membrane plays in the interaction between ligands and transmembrane proteins. Gating modifier toxins (GMTs), a class of disulfide-rich peptides derived from spider venom, have proven to be useful tools in the study of voltage-gated ion channels, which are transmembrane proteins integral to several (patho)physiological processes. Nine GMTs were chemically synthesized and a range of biophysical techniques was used to evaluate the structures and physicochemical properties of the peptides that contribute to interactions with lipid membranes. The GMTs were then assessed for activity at voltage-gated sodium channel subtype 1.7 (NaV1.7), which mediates pain in humans. We observed that for some GMTs, potent inhibition of NaV1.7 is accompanied with binding to lipid membranes, and that the presence of an electronegative charge on the solvent exposed surface of some GMTs can hinder interactions with lipid membranes. This information guided the design of gHwTx-IV, a GMT analogue with a large electropositive solvent exposed surface. gHwTx-IV showed increased affinity for the lipid membrane and was a more potent inhibitor of NaV1.7, compared to HwTx-IV, the parent peptide. The interactions between gHwTx-IV and the cell membrane are predominantly electrostatic, and occur at the water/lipid interface, indicating that the membrane acts to attract and anchor GMTs near their NaV1.7 active site to optimize inhibition. The present work is novel evidence that the membrane adds a third dimension to the lock-in-key approach of rational drug design.