The objective was to compare three score systems, pneumonia severity index (PSI), the Confusion-Urea-Respiratory Rate-Blood pressure-65 (CURB-65), and severe community-acquired pneumonia (SCAP), for prediction of the outcomes in a cohort of patients with community-acquired (CAP) and healthcare-associated pneumonia (HCAP). Large multi-center, prospective, observational study was conducted in 55 hospitals. HCAP patients were included in the high classes of CURB-65, PSI and SCAP scores have a mortality rate higher than that of CAP patients. HCAP patients included in the low class of the three severity rules have a significantly higher incidence of adverse events, including development of septic shock, transfer into an ICU, and death (p < 0.01). At multivariate Cox regression analysis, inclusion in the severe classes of PSI, CURB-65, or SCAP scores and receipt of an empirical therapy not adherent to international guidelines prove to be risk factors independently associated with poor outcome. PSI, CURB-65, and SCAP score have a good performance in patients with CAP but are less useful in patients with HCAP, especially in patients classified in the low-risk classes.