OBJECTIVE To assess the relationship between concentrations of bronchoalveolar cytokines and bacterial burden (quantitative bacterial count) in intubated patients with a presumptive diagnosis of community-acquired pneumonia. DESIGN A cross-sectional and clinical investigation. SETTING Medical/surgical and respiratory intensive care unit of a tertiary 1,200-bed medical center. PATIENTS According to the time course of community-acquired pneumonia at the time of study with bronchoalveolar lavage, 69 mechanically ventilated patients were divided into three subgroups: primary (n = 11), referral (n = 23), and treated (n = 35) community-acquired pneumonia. INTERVENTIONS Bronchoalveolar lavage was performed in the most abnormal area on chest radiograph by fiberoptic bronchoscope. Bronchoalveolar lavage fluid was processed for quantitative bacterial culture. The concentrations of bronchoalveolar lavage cytokines (tumor necrosis factor-alpha, interleukin-1 beta, interleukin-6, interleukin-8, and interleukin-10) also were measured. MEASUREMENTS AND MAIN RESULTS Thirty-two patients had a positive bacterial culture (bronchoalveolar lavage > or = 10 colony-forming units/mL)., and made up 76% of pathogens recovered at high concentrations. The concentrations of bronchoalveolar lavage interleukin-1 beta were 199.1 +/- 32.1 and 54.9 +/- 13.0 pg/mL (mean +/- se) in the patients with positive and negative bacterial culture, respectively (p < .001). Bronchoalveolar lavage interleukin- 1 beta was significantly higher in the patients with a high bacterial burden (p < .001), with mixed bacterial infection (p < .001), and with pneumonia (p < .001), compared with values in patients without these features. The relationship between bacterial load and concentrations of bronchoalveolar lavage interleukin-1 beta was very strong in the patients with primary and referral community-acquired pneumonia but was borderline in treated community-acquired pneumonia. CONCLUSIONS The common pathogens were similar to the core pathogens of hospital-acquired pneumonia, probably due to antibiotic effects, delayed sampling, and superimposed nosocomial infection. Since the concentration of bronchoalveolar lavage interleukin-1 beta was correlated with bacterial burden in the alveoli, it may be a marker for progressive and ongoing inflammation in patients who have not responded to pneumonia therapy and who have persistence of bacteria in the lung.