We have modeled platelet aggregation in a linear shear flow by accounting for two body collision hydrodynamics, platelet activation and receptor biology. Considering platelets and their aggregates as unequal-sized spheres with DLVO interactions (psi(platelet) = -15 mV, Hamaker constant = 10(-19) J), detailed hydrodynamics provided the flow field around the colliding platelets. Trajectory calculations were performed to obtain the far upstream cross-sectional area and the particle flux through this area provided the collision frequency. Only a fraction of platelets brought together by a shearing fluid flow were held together if successfully bound by fibrinogen cross-bridging GPIIb/IIIa receptors on the platelet surfaces. This fraction was calculated by modeling receptor-mediated aggregation using the formalism of Bell (Bell, G. I. 1979. A theoretical model for adhesion between cells mediated by multivalent ligands. Cell Biophys. 1:133-147) where the forward rate of bond formation dictated aggregation during collision and was estimated from the diffusional limited rate of lateral association of receptors multiplied by an effectiveness factor, eta, to give an apparent rate. For a value of eta = 0.0178, we calculated the overall efficiency (including both receptor binding and hydrodynamics effects) for equal-sized platelets with 50,000 receptors/platelet to be 0.206 for G = 41.9 s(-1), 0.05 for G = 335 s(-1), and 0.0086 for G = 1920 s(-1), values which are in agreement with efficiencies determined from initial platelet singlet consumption rates in flow through a tube. From our analysis, we predict that bond formation proceeds at a rate of approximately 0.1925 bonds/microm2 per ms, which is approximately 50-fold slower than the diffusion limited rate of association. This value of eta is also consistent with a colloidal stability of unactivated platelets at low shear rates. Fibrinogen was calculated to mediate aggregation quite efficiently at low shear rates but not at high shear rates. Although secondary collisions (an orbitlike trajectory) form only a small fraction of the total number of collisions, they become important at high shear rates (>750 s(-1)), as these are the only collisions that provide enough time to result in successful aggregate formation mediated by fibrinogen. The overall method provides a hydrodynamic and receptor correction of the Smoluchowski collision kernel and gives a first estimate of eta for the fibrinogen-GPIIb/IIIa cross-bridging of platelets. We also predict that secondary collisions extend the shear rate range at which fibrinogen can mediate successful aggregation.