Norwalk Virus (NV) and other human enteric viruses in fecally contaminated waters pose a human health risk that is inadequately characterized due to the limitations of current virus detection methods. Standard methods for virus concentration from water using electropositive adsorbent filters have not been tested for NV recovery because there is no convenient infectivity assay system and because the beef extract (BE/G) used to elute the adsorbed viruses inhibits virus detection by reverse transcriptase-polymerase chain reaction (RT-PCR) amplification. Hence, we tested five alternative eluents compatible with RT-PCR (the amino acids arginine, asparagine, glycine, lysine, orthreonine) for elution of NV as well as polio virus type 1 (PVl) and coliphage MS2 (MS2). All three viruses seeded in to dechlorinated tap water were adsorbed to Virosorb 1MDS filters and then eluted with either BE/G or an alternative eluent to achieve a 100-fold volume concentration. All three viruses were assayed by endpoint dilution RT-PCR and PVl and MS2 were assayed by infectivity using the plaque technique as well. With all of the amino acids except threonine, elution efficiency for all three viruses was comparable to or greater than that by BE/G. Virus recovery efficiencies were 50-75% over a pH range of 7.0-9.8. Viruses were concentrated an other 10-fold by polyethylene glycol precipitation, which was also compatible with molecular detection techniques and gave recovery efficiencies of 10-100%. These results indicate that modified filter adsorption-elution methods using amino acid eluents and subsequent secondary concentration by polyethylene glycol precipitation provide efficient concentration of NV and other viruses in water for subsequent assay by RT-PCR.