The objectives of the present study were to: (1) evaluate the safety of Filgrastim therapy in nonneutropenic patients with severe community-acquired pneumonia; (2) determine the absolute neutrophil count (ANC) response to various dosages of Filgrastim in non-neutropenic patients with active infection; and (3) describe the impact of therapy with Filgrastim in combination with antibiotics on selected pneumonia-related clinical parameters. The study design was an open-label, dose-ranging, clinical trial, set in the General Clinical Research Unit of a large, public community hospital. The study population consisted of 30 patients who had presented to the Emergency Department with severe, community-acquired pneumonia. One of five dosages (75, 150, 300, 450 or 600 mu g day(-1)) of Filgrastim (r-metHuG-CSF) was given subcutaneously daily for 10 days, until discharge or until the absolute neutrophil count >75 x 10(9) l(-1), whichever was earlier. Vital signs, pulse oximetry, arterial blood gases, daily complete blood counts with differential, serum chemistries, coagulation profiles, electrocardiograms, chest radiographs, plasma G-CSF concentrations and duration of hospitalization were measured. There was no evidence of Filgrastim-related lung injury or evidence of extra-pulmonary toxicity. There was no apparent dose-response effect of Filgrastim on pneumonia-related clinical variables. Dosages of Filgrastim between 150 and 600 mu g day(-1) had similar effects on increasing the ANC. Filgrastim appeared to be safe in non-neutropenic patients with severe, community-acquired pneumonia when given in dosages of 75-600 mu g day(-1) in combination with appropriate antibiotic therapy. Further study is needed to determine the effect of Filgrastim on morbidity, mortality and duration of symptoms in this patient population.