In 1987 we reported that when blood was forced through a fine filter under pressure in the filterometer the platelets aggregated and blocked the filter, von Willebrand factor (vWF) and glycoprotein (Gp) Hb/IIIa and calcium were involved. Results with anti-GpIb were equivocal. We now report that all the anti-GpIb antibodies studied, glycocalicin, as well as some concentrations of aurin tricarboxylic acid caused platelet aggregation in the pre-filter blood and therefore could not be used in the filterometer. Using two different molecules that prevent vWF binding to GpIb and two anti-GpIIb/IIIa antibodies at two pressures it has now been shown that GpIb, vWf and high shear are primarily responsible for platelet retention at 0–5 s. Progressive platelet retention studied between 20 and 40 s required high shear and GpIIb/IIIa after the calcium influx mediated by GpIb/vWF binding. When GpIb was inhibited, GpIIb/IIIa could not function normally, so GpIb inhibition resulted in decreased aggregation both at 0–5 s and at 20–40 s. Anti-GpIIb/IIIa caused a minimal decrease in retention at 0–5 s and marked inhibition at 20–40 s. These findings fit and amplify concepts derived from other high shear methodologies. A diagram is presented of the events leading up to the final 'passivation' of the 'thrombus' in the filter when the surface of the aggregated platelets becomes unattractive.