Background: The differential diagnosis of community-acquired pneumonia and some non-pneumonia diseases involving the chest may sometimes be cumbersome. Adding some objective variables to the diagnostic strategy may be helpful.We evaluated the main objective variables that are usually available in the emergency ward and that may be valuable in this differential diagnosis. Methods: We recorded epidemiological, clinical, and analytical data, as well as that obtained from physical examination, from 284 consecutive patients diagnosed in the emergency ward as having community-acquired pneumonia. The diagnosis was reviewed by the investigators applying pre-set diagnostic criteria. Statistical analysis was then performed comparing data from patients with a definitive diagnosis of community-acquired pneumonia with those with a final diagnosis of non-pneumonia disease excluding acute exacerbations of chronic bronchitis. Results: In the univariate analysis, C-reactive protein (difference of means 93 mg/l; 95% C.I. 47, 140), erythrocyte sedimentation rate (d.m. 19 mm/h; 95% C.I. 3, 35), leukocyte count (d.m. 3.5x10(9)/l; 95% C.I. 0.5, 6.4), and temperature (d.m. 0.5 degrees C; 95% C.I. 0.1, 0.9) discriminated between community-acquired pneumonia and non-pneumonia diseases. In the multivariate analysis, only C-reactive protein remained in the equation. Conclusions: C-reactive protein, erythrocyte sedimentation rate, leukocyte count, and temperature were measurable variables that proved to be useful in the differential diagnosis between community-acquired pneumonia and non-pneumonia diseases. C-reactive protein appears to be the most suitable for this purpose.