The purpose of this study was to develop methods for detection and quantitation of food-borne virus. Samples (25 g) of cottage cheese, contaminated with various quantities of coxsackievirus type A9, comprised the model system. Two of the methods presented have at least a 50% probability of detecting virus at levels below 5 plaque-forming units/25-g sample. Noteworthy aspects of these methods include use of a glycine-NaOH buffer (pH 8.8) containing approximately 1 m MgCl(2) as the diluent in which the sample is slurried, treatment of the slurry with Freon TF and bentonite to facilitate centrifuge clarification, and concentration of the clarified sample extract by a two-stage process employing polyethylene glycol followed by ultracentrifugation. Virus in the final 0.5-ml sample concentrate was detected and quantitated by the plaque technique in rhesus monkey kidney cell cultures. Processing of the sample requires approximately 2 days, and the inoculated cultures may have to be observed for as long as 7 days thereafter. If these levels of sensitivity are desired, and if 12 samples per day are tested on a routine basis, the cost savings achieved by employing these methods rather than testing sample extracts without concentration may range from 75 to 90%.