NEUROTOXINS that selectively block Na+, K+ or Ca2+ channels have provided valuable information about the functional diversity of the voltage-gated channel superfamily(1). For Ca2+ channels, a variety of toxins have been found to block individual channel types(2,3). The best-known example is omega-conotoxin-GVIA, a member of a large family of peptide toxins derived from venomous cone snails(2,3), which potently and selectively blocks N-type Ca2+ channels(4-9), allowing their purification(10,11), cellular localization(12,13), and the elucidation of their roles in Ca2+ entry(14), neurotransmitter release(15,16) and neuronal migration(17). In contrast to Na+ and K+ channels, little is known about the molecular features that underlie Ca2+-channel susceptibility to toxin block; it is also unknown whether block occurs by direct physical occlusion(3) or an action on channel gating(18). Here we describe structural determinants of the N-type Ca2+ channel’s interaction with omega-conotoxin-GVIA. When chimaeras combining individual motifs from the N-type channel and from a channel insensitive to omega-conotoxin-GVIA were expressed in Xenopus oocytes, each of the four motifs appeared to contribute to interaction with the toxin. The most dramatic effects on toxin interactions were seen at a single cluster of residues in the large putative extracellular loop between IIIS5 and IIIH5, consistent with a direct pore-blocking mechanism. These results provide a starting point for delineating the architecture of the outer vestibule of the Ca2+ channel.