Intrinsic fluorescence of myotoxin a from Crotalus viridis viridis has been proved to be due to tryptophan residues, and this knowledge allows one to detect, for the first time, an interaction of the toxin with phospholipids. The formation of lipid-protein species is followed by a blue shift of about 8 nm and a quenching of up to 50% of the emission of tryptophans. These changes could result from the positioning of Trp 32,34 residues at the lipid interface and/or a local conformational change of the toxin. In the μM range, complexes are formed only with charged lipids; charge complementarity of two partners is needed regardless of the physical state of the lipids. The binding is electrostatic, and the affinity of myotoxin a decreases in parallel with the net charge of the phospholipid interface and/or that of the toxin. Similarly, calcium-ion addition, pH changes, and an increase in ionic strength can dissociate the complexes. Their stability increases in the following sequence: phosphatidylinositol < phosphatidylserine < phosphatidylglycerol < phosphatidic acid < cardiolipin. It is proposed that the toxin lies at the interface, without penetrating the membrane. On binding, the toxin does not severely change the structure of lipid vesicles nor their permeability; however, significant changes in the lipid chain order are detected with synthetic lipids, both in their gel and fluid phases. The possible consequences of the existence of myotoxin a-charged phospholipid complexes are discussed and are compared to the behavior of other related basic toxins of snake venom acting on membrane.