A simple model system has been developed in which lectin-mediated aggregation of glycoprotein-coated beads can be monitored by following the decrease in light scattering at 650 nm. Aggregation has been characterized with the lectin of Actinomyces viscosus T14V. Its dependence on pH, temperature, and stirring rate was examined, and the number of bacterial cells in relation to the number of latex beads resulting in optimal aggregation was established. This system has the advantage of permitting the study of a single ligand of defined structure. The ligand density was determined with radiolabeled glycoproteins. Under the conditions of the assay, ligand leakage was less than 3%, and ligands were not displaced from the beads by various proteins, glycoproteins, or by other components present in the assay mixture. Latex beads coated with asialofetuin aggregate upon the addition of A. viscosus T14V cells. By contrast, when asialofetuin was first extensively treated with purified galactose oxidase, no aggregation occurred. Only after reduction with NaBH 4 was aggregation restored, demonstrating that galactose termini of asialofetuin are essential for the binding of A. viscosus lectin. An absolute requirement for calcium was also demonstrated. Various sugars inhibited aggregation in the following order, starting with the most effective: lactose, methyl-β-D-galactopyranoside, galactose, N -acetylgalactosamine, methyl-α-D-galactopyranoside. Beads coated with fimbriae from A. viscosus coaggregated with neuraminidase-treated human erythrocytes and with Streptococcus sanguis cells. In each instance the aggregation was inhibited by lactose, indicating that the A. viscosus lectin is located in the fimbriae. Cells grown under different conditions differed in their effectiveness in aggregating glycoprotein-coated beads, suggesting differences in lectin density or accessibility. Two different experimental designs were used to establish the minimum ligand density for aggregation to occur. In one type of experiment, a threshold concentration was found for asialo α 1 -acid glycoprotein, but not for asialofetuin. With an alternate approach in which a different population of galactose residues was exposed, a threshold phenomenon was also demonstrated for asialofetuin. The importance of structural ligand features in the aggregation assay is discussed in view of these findings.