Summary Objective To identify modifiable risk factors of nosocomial pneumonia (NP) in elderly patients post-cardiac surgery. Design A case-control study. Setting Post-operative intensive care unit of a tertiary-level university affiliated hospital. Subjects Seventy three case-control pairs. Case patients referred to elderly patients who developed pneumonia post-cardiac surgery. Controls subjects were matched for age, gender, type of surgery, forced expiratory volume in 1s (FEV 1 ), and ejection fraction. Measurements Baseline sociodemograpahic information, Charlson Comorbidity Index score, intra- and post-operative data were collected. When suspected, the presence of NP was confirmed by quantitative culture of protected bronchoalveolar lavage fluid ⩾10 3 colony forming unit/ml or positive blood/pleural fluid culture identical to that recovered from respiratory samples. Results The incidence of NP in elderly post-heart surgery was 8.3%. The mean duration after heart surgery to the occurrence of pneumonia was 7.2±4.9 days. Four variables were found to be significantly related to the development of NP by multivariate analysis: Charlson Index >2 (adjusted odds ratio [AOR] 4.7; 95% confidence interval [CI], 1.9–11.4; P 0.001 ), reintubation (AOR 6.2; 95% CI, 1.1–36.1; P = 0.04 ), transfusion ⩾4 units of PRBC (AOR 2.8; 95% CI, 1.2–6.3; P = 0.01 ), and the mean equivalent daily dose of morphine (AOR 4.6; 95% CI, 1.4–14.6; P = 0.01 ). Conclusions Although there are limited effective measures to lessen the burden of comorbidities, avoiding reintubation, finding a substitute to allogenic blood transfusion, and improved assessment of pain management could reduce the rate of NP in the post-operative period of cardiac surgery in the elderly population.