In acidic solution, thorin (2-(2-hydroxy-3.6-disulfo-1-naph-thy1)-azo-phenylarsenic acid sodium salt) can be bound with protein and aggregated to form large particle which displays a very strong rayleigh light scattering (RLS). The effects of pH, thorin concentration, detergent and ionic strength on binding reaction have been studied. The interference of coexisting substances was checked. The Scatchard plots for reaction between thorin and bovine serum albumin (BSA) were constructed and the association constant of thorin-BSA was obtained, it is 5.26×10 5 L/mol, the maximum binding number is 7. RLS intensity is well proportional to the concentrations of 2.0-14.0 μg/mL for human serum albumin (HSA), 1.8-14.7 μg/mL for BSA and 1.8-14.6 μg/mL for γ-globulin (γ-G). The detection limits (3σ) are 54.1 ng/mL for HSA, 52.0 ng/mL for BSA and 51.8 ng/mL for γ-G, respectively. The relative standard deviation is 2.4% for BSA, 3.2% for HSA and 4.1% for γ-G. The human serum samples were measured satisfactorily by using this method. In acidic solution, thorin (2- (2-hydroxy-3.6-disulfo-1-naph-thy1) -azo-phenylarsenic acid sodium salt can be bound with protein and aggregated to form large particle which displays a very strong rayleigh light scattering The effects of pH, thorin concentration, detergent and ionic strength on binding reaction have been studied. The interference of coexisting substances was checked. The Scatchard plots for reaction between thorin and bovine serum albumin (BSA) were constructed and the association The maximum binding number is 7. RLS intensity is well proportional to the concentrations of 2.0-14.0 μg / mL for human serum albumin (HSA), 1.8- 14.7 μg / mL for BSA and 1.8-14.6 μg / mL for γ-globulin (γ-G). The detection limits (3σ) are 54.1 ng / mL for HSA, 52.0 ng / mL for BSA and 51.8 ng / mL for γ -G, respectively. The relative standard deviation is 2.4% for BSA, 3.2% for HSA and 4.1% for γ-G. The human serum samples were me asured satisfactorily by using this method.