Influenza infection results in substantial morbidity and mortality in hospitalized patients, including those who are immunocompromised or pregnant. Antiviral therapy likely provides considerable benefit to these patients, but few studies have been successfully conducted in these high-risk populations, and no drugs are specifically licensed for treating these subgroups. One of the key challenges facing novel antiviral drug development for influenza is determining the appropriate efficacy end points that would enable rapid regulatory approval for drug use in seriously ill patients, for whom risk-benefit assessments differ from those with uncomplicated illness. All available antiviral drugs currently affect viral replication, and respiratory tract viral titers correlate with both symptoms and measures of host inflammatory responses, including cytokine and chemokine expression that are likely responsible for many of the clinical symptoms. Consequently, we outline the evidence to support the use of primary virological end points in studies of antiviral agents involving patients who are hospitalized with severe influenza or those who are at high risk of severe and life-threatening disease.