Granulocyte colony stimulating factor (G-CSF) is a naturally-occurring cytokine that has been shown to increase neutrophil function and number. Exogenous administration of recombinant G-CSF (filgrastim, pegfilgrastim or lenograstim) has found extensive use in the treatment of febrile neutropaenia, but its role in the treatment of infection in non-neutropaenic hosts is less well defined. We aimed to explore the role of G-CSF as an adjunct to antibiotics in the treatment of pneumonia in non-neutropaenic adults. A search was performed using the Cochrane Central Register of Controlled Trials (issue 1, 2003); MEDLINE (January 1966 to April 2003); EMBASE (1988 to 2003); online databases of clinical trials; contact with corresponding authors; and contact with the manufacturers and distributors of G-CSF and reviews of citations in publications identified by the above strategies. We considered randomised controlled trials (RCTs) which included hospitalised adult patients with either community acquired pneumonia or hospital-acquired pneumonia. Studies identified were reviewed independently by two reviewers with data abstracted onto standardized data collection forms. The primary outcome measure was 28 day mortality. Secondary outcome measures included other markers of mortality as well as markers of adverse events, including organ dysfunction. An assessment of methodological quality was made for each study. G-CSF use appeared to be safe with no increase in the incidence of total serious adverse events (pooled OR 0.91, 95% CI: 0.73, 1.14) or organ dysfunction. However, the use of G-CSF was not associated with improved 28 day mortality (pooled OR 0.99, 95% CI 0.77, 1.29). There is no current evidence supporting the routine use of G-CSF in the treatment of pneumonia. Studies in which G-CSF is administered prophylactically or earlier in therapy may be of interest.