Introduction: To identify factors influencing decisions in initial management of community-acquired pneumonia (CAP) admitted to hospital through Emergency departments., Methods: Records of CAP adult patients admitted to 24 Spanish hospitals in January-March 2003 were reviewed. Patients sent for ambulatory treatment were excluded., Results: 341 patients (67.0 +/- 24.6 years; 65.3% males) were included; 39% were taking antibiotics at attendance. PSI was (% patients): I-II (19.7%), III (14.7%), and IV-V (65.6%). Comorbidities were: COPD (37.2%), heart disease (24.6%), hypertension (17%), diabetes mellitus (10.8%), and malignancies (10%). Pneumococcal/Legionella urinary antigens were performed in 34.0%/42.2% patients. Fewer (p < or = 0.006) rapid tests were performed in class IV-V (p = 0.001), with higher (p < or = 0.01) pneumococcal positive results in class V. Initial treatment was fluoroquinolone (37.5%), beta-lactam + macrolide (26.4%), beta-lactam (22.9%), macrolide (4.7%), and others (8.5%). Patients referred to Internal Medicine had higher heart disease (p = 0.06) and hypertension (p = 0.001) as comorbidity than those at Short-Stay Units or Pneumology. COPD patients were equally distributed between Internal Medicine and Pneumology, with differences vs. Short-Stay Units., Conclusions: Rapid diagnostic tests were underused, maybe due to broad empirical treatments covering drug-resistant pneumococci and L. pneumophila (regardless PSI and comorbidity). Presence of comorbidities or positive results in rapid diagnostic tests seems to influence the medical ward to which the patient is referred to, but not initial treatment.