Background. Readmission rates following discharge after pneumonia are thought to represent the quality of care. Factors associated with readmission, however, remain poorly described. It is unclear if readmission rates vary based on pneumonia type. Methods. We retrospectively identified adults admitted to an index hospital with non-nosocomial pneumonia ( January through December 2010) and who survived to discharge. We only included patients with bacterial evidence of infection. Readmission in the 30 days following discharge to any of 9 hospitals comprising the index hospital's healthcare system served as the primary end point. We recorded demographics, severity of illness, comorbidities, and infection-related factors. We noted whether the patient had healthcare-associated pneumonia (HCAP) versus community-acquired pneumonia. We utilized logistic regression analysis to determine factors independently associated with readmission. Results. The cohort included 977 subjects; 78.9% survived to discharge. The readmission rate equaled 20%. Neither disease severity nor the rate of initially inappropriate antibiotic therapy correlated with readmission. Subjects with HCAP were 7.5 (95% confidence interval [CI], 3.6-15.7) times more likely to be readmitted. Four HCAP criteria were independently associated with readmission: admission from long-term care (adjusted odds ratio [AOR], 2.2 [95% CI, 1.4-3.4]); immunosuppression (AOR, 1.9 [95% CI, 1.3-2.9]); prior antibiotics (AOR, 1.7 [95% CI, 1.2-2.6]); and prior hospitalization (AOR, 1.7 [95% CI, 1.1-2.5]). Conclusions. Readmission for pneumonia is common but varies based on pneumonia type. The variables associated with readmission do not reflect factors that hospitals directly control. Use of one rule to guide payment that fails to account for HCAP and the HCAP criteria on readmission seems inappropriate.