Severe combined immunodeficient (SCID) mice have been evaluated for applicability as hosts for a human immunodeficiency virus (HIV) animal model, compatible with the pathogenesis of HIV disease and/or for testing compounds for antiviral efficacy. McCune et al. [Science 241 (1988) 1632] described the SCID/hu model and Namikawa et al. [J. Exp. Med. 172 (1990) 1055] and Rabin et al. [Antimicrob. Agents Chemother. 40 (1996) 755] described the SCID/hu (Thy/Liv) model which was developed for the evaluation of HIV pathogenic mechanisms and for the prioritization of antiviral compounds that were efficacious in vitro. Hollingshead et al. [Antiviral Res. 28 (1995) 265] and Xu et al. [Bioorg. Med Chem. Lett. 9 (1999) 133] described the HIV hollow fiber SCID mouse model. This model was developed to be a low cost, high throughput, time efficient, simple in vivo screening system for preliminary anti-HIV efficacy evaluation for the prioritization of antiviral compounds that demonstrated in vitro efficacy. The hollow fiber model is used as a pharmacologic tool to help separate active and inactive agents and direct the best lead compounds into additional animal model testing (e.g. SCID/hu). Compounds that are known to have an antiviral effect in man (e.g. 3'-azo-3'-deoxythymidine (AZT), dideoxyinosine (ddI) and dideoxycytidine (ddC)) were evaluated in both models. The endpoints (e.g. PCR, flow cytometry, MTT, p24, RT) evaluated in both models indicate that HIV-1 virus replicates in both models and infection is suppressed in the SCID/hu and hollow fiber SCID mouse models when treated with approved clinical antiviral agents. While both models are useful for the evaluation of antiviral therapies, there are distinct advantages (e.g. cost, time, material, equipment, expediency) with the hollow fiber assay over the SCID/hu model (Thy/Liv) for antiviral drug evaluations particularly in terms of cost effectiveness.