The major lethal component of the venom of Naja naja philippinensis has been extracted and extensively purified, up to complete homogeneity, by physical and chemical standards and by exacting immunological tests. The method of preparation, aimed at preserving the native conformation, avoids the polystyrene resins commonly used for such toxins, but relies exclusively on hydrophilic supporting media and takes advantage of the extreme affinity of salt-free neurotoxin for wetted poly-dextran gels in order to increase the efficiency considerably. The pure toxin has an LD50of 0.05 μg/g, is a single chain polypeptide of 61 residues, has a mol. wt. of 6867, a pI of 10.8, and an unusually high absorptivity, 14,300 at 280 nm, when compared to all known parent toxins. At 10-4to 10-5M and slightly acidic pH, equilibrium sedimentation failed to provide evidence for dimerization or conformational heterogeneity. This minute and tightly assembled protein molecule belongs to type I neurotoxins of snake venoms, acting selectively by inhibiting myofibrillar membrane depolarization in excitatory synapses and blocking myoneural transmission to striated muscles, particularly those involved in pulmonary ventilation. This neurotoxin brings about two new and unexpected features: 1) It differs from all known isotoxins of type 1 by having Ala in its composition, located at position 11 in the sequence, an amino acid so far considered as typical of type II neurotoxins of more recent origin, and of all other isotoxins of higher mol. wt. 2) Strikingly, it contains two Trp instead of the unique Trp characteristic of all neurotoxins described so far. Though the unique ancestral Trp has been claimed repeatedly to play a critical role in lethality of all neurotoxins, the present sidewise duplication shows no influence on the LD50. ORD-CD profiles display β sheet structure predominantly, an observation consistent with the occurrence of a majority of non helix-formers in the two largest stretches. A third large loop, where hydrophobic and all the aromatic side chains are packed, may have helical conformation. Though denaturation in 6 M guanidine-HCl releases completely randomized structures, as apparent by ORD-CD, the unfolding in 10 M urea is far from complete and remains fully reversible. Fluorescence investigations by reversible solvent perturbation using methanol or urea confirm localized unfolding accompanied by the translation of the unique Tyr chromophore from the interior out to the surface of the molecule, whereas both Trp chromophores appear unaffected except for minor hyperchromatism. These residues may not necessarily be directly involved in the receptor-complex formation, but may be essential for conformational stabilization. Whether neurotoxin engages in this binding process as a rigid ligand or by some kind of “induced-fit” is a matter for further investigations.