It was discovered that the product of a mix containing the enzyme creatine phosphokinase (CPK) and either glutathione (GSH) or cysteine caused platelets to adhere together in vitro. This platelet adhesive factor (PAF) was formed as CPK enzyme activity declined. An alternative method for the destruction of enzyme activity--heat at 56 degrees C--also resulted in the formation of an in-vitro active PAF which was both less stable and active than its chemically produced counterpart. Assay of the platelet adhesive potency of the CPK-GSH mix, using human platelets, revealed a wide variation in the response of different individuals' platelets to standard quantities of PAF. The nature of this preparation of PAF was investigated by both biochemical and biophysical means, including ion exchange chromatography, electrophoresis, amino acid analysis and analytical ultracentrifuge studies. Evidence is presented that PAF is the product of the disruption of the dimeric structure of the CPK molecule. PAF was found to adhere to paper, under the conditions of electrophoresis imposed, and also to cause sephadex beads to bind together, characteristics which suggested that the platelet adhesion reaction was probably a biophysical process. Red and white cells were not similarly affected. The feasibility of this novel concept for the initiation of platelet adhesion, as a naturally occurring process, was supported by the results of animal experiments in which a statistically depression of platelets in the systemic circulation followed the intravascular administration of PAF. The possible relevance to man of this basic mechanism in relation to exercise and disease processes, including ideopathic and post-traumatic thrombosis, atherogenesis, and dysbaric aseptic necrosis of bone, is discussed.