Patients with acute, diffuse pulmonary infiltration that may be infectious or noninfectious present a major diagnostic-treatment challenge. Many options are available during the sequence of diagnosis and treatment. Our bias is to proceed rapidly to the more specific and sensitive invasive diagnostic procedures if clinical and laboratory data are indicative of diffuse involvement or immunosuppression, because a positive outcome appears to be affected by early intervention with appropriate antimicrobial drugs. Such an approach also avoids administration of a potentially toxic and unnecessary antimicrobial "shotgun." Each physician must consider the local expertise and resources available; in our view, however, the patient is best served by, "erring" on the side of early use of invasive procedures.