Antibiotic combinations that include macrolides have shown lower mortality rates than β-lactams in monotherapy or combined with fluoroquinolones in patients with community-acquired pneumonia (CAP). However, this effect has not been studied according to the levels of C-reactive protein in CAP with identified microbial cause. In patients with CAP and known microbial cause we aimed to evaluate 30-day mortality of a β-lactam plus macrolide (BL + M) compared with a fluoroquinolone alone or with a β-lactam (FQ ± BL). METHODS: We analyzed a prospective observational cohort of patients with CAP admitted to the Hospital Clinic of Barcelona between 1996 and 2016. We included only patients with known microbial cause. RESULTS: Of 1,715 patients (29%) with known etiology, a total of 932 patients (54%) received BL + M. Despite lower crude mortality in the BL + M group in the overall population (BL + M, 5% vs FQ ± BL, 8%; P = .015), after adjustment by a propensity score and baseline characteristics, the combination of BL + M had a protective effect on mortality only in patients with high inflammatory response (C-reactive protein, > 15 mg/dL) and pneumococcal CAP (adjusted OR, 0.28; 95% CI, 0.09-0.93). No benefits on mortality were observed for the population without high inflammatory response and pneumococcal CAP or with other etiologies. CONCLUSIONS: The combination of a β-lactam with a macrolide was associated with decreased mortality in patients with pneumococcal CAP and in patients with high systemic inflammatory response. When both factors occurred together, BL + M was protective for mortality in the multivariate analysis.