The need for rapid, highly sensitive, and versatile diagnostic tests for viral pathogens spans from human and veterinary medicine to bioterrorism prevention. As an approach to meet these demands, a diagnostic test employing monoclonal antibodies (mAbs) for the selective extraction of viral pathogens from a sample in a chip-scale, sandwich immunoassay format has been developed using surface-enhanced Raman scattering (SERS) as a readout method. The strengths of SERS-based detection include its inherent high sensitivity and facility for multiplexing. The capability of this approach is demonstrated by the capture of feline calicivirus (FCV) from cell culture media that is exposed to a gold substrate modified with a covalently immobilized layer of anti-FCV mAbs. The surface-bound FCVs are subsequently coupled with an extrinsic Raman label (ERL) for identification and quantification. The ERLs consist of 60-nm gold nanoparticles coated first with a layer of Raman reporter molecules and then a layer of mAbs. The Raman reporter molecule is strategically designed to chemisorb as a thiolate adlayer on the gold nanoparticle, to provide a strong and unique spectral signature, and to covalently link a layer of mAbs to the gold nanoparticle. The last feature provides a means to selectively tag substrate-bound FCV. This paper describes the development of the assay, which uses cell culture media as a sample matrix and has a linear dynamic range of 1 x 10(6)-2.5 x 10(8) viruses/mL and a limit of detection of 1 x 10(6) viruses/mL. These results reflect the findings from a detailed series of investigations on the effects of several experimental parameters (e.g., salt concentration, ERL binding buffer, and sample agitation), all of which were aimed at minimizing nonspecific binding and maximizing FCV binding efficiency. The performance of the assay is correlated with the number of captured FCV, determined by atomic force microscopy, as a means of method validation.