Hydrolysis of gel phase dipalmitoylphosphatidylcholine (DPPC) at 37 degrees C catalysed by Crotalus atrox phospholipase A2 (PLA) is described extremely well by the "path 1" kinetic mechanism of Tinker and Wei (1979) (Can. J. Biochem. 57, 97-106), if reversible adsorption is allowed as a side reaction. Progress curves show an initial rapid phase, the initial velocity being a Michaelis-Menten function dependent on the catalytic properties of the enzyme (kcat approximately equal to 9200 min-1, Km approximately equal to 0.12 mM), then level off to a slower rate determined by the desorption equilibrium constant (KD approximately equal to 0.01 mM) and desorption rate constant (kD approximately equal to 0.15 min-1). The relaxation time, tau, for the transition to the desorption-limited reaction is approximately 0.5 min; this large value of tau probably arises from a slow conversion of active, dimeric enzyme to an inactive protein species adsorbed to the lipid surface. At later times in the reaction there is an increase in the rate of hydrolysis, attributed to a stimulation of desorption by the products. The desorption equilibrium constant KD is a quadratic function of the surface concentration of products and increases 20- to 30-fold when all accessible substrate is hydrolysed. Both lysophosphatidylcholine (lyso-PC) and fatty acid were found to stimulate the desorption, but lyso-PC was also found to be a competitive inhibitor of the catalysis. Adsorption of PLA to DPPC and egg PC vesicles was directly measured using a gel partition technique. Strong binding to egg PC was observed, which was not dependent on the presence of calcium ion (essential for catalysis); PLA inhibited by acylation of up to four lysine residues per mole of monomeric enzyme with ethoxyformic anhydride was equally strongly adsorbed, indicating that lipid binding is not dependent on catalytic activity. Reaction products greatly weakened the binding of PLA to the lipid surface as expected. Cholesterol had two effects on the hydrolytic reaction: there was a striking decrease in the rate of the slower, desorption-limited phase, the rate of which decrease to almost zero at 15 mol% cholesterol, but there was also evidence for the formation of a complex with stoichiometry 1 cholesterol: 2 DPPC in which DPPC is no longer a substrate for the enzyme. Implications of the proposed mechanism for specificity and control of surface catalysis by PLA are discussed.