To compare efficacy and safety of a tigecycline regimen with an imipenem/cilastatin regimen in hospital-acquired pneumonia patients, a phase 3, multicenter, randomized, double-blind, study evaluated 945 patients. Coprimary end points were clinical response in clinically evaluable (CE) and clinical modified intent-to-treat (c-mITT) populations at test-of-cure. Cure rates were 67.9% for tigecycline and 78.2% for imipenem (CE patients) and 62.7% and 67.6% (c-mITT patients), respectively. A statistical interaction occurred between ventilator-associated pneumonia (VAP) and non-VAP subgroups, with significantly lower cure rates in tigecycline VAP patients compared to imipenem; in non-VAP patients, tigecycline was noninferior to imipenem. Overall mortality did not differ between the tigecycline (14.1%) and imipenem regimens (12.2%), although more deaths occurred in VAP patients treated with tigecycline than imipenem. Overall, the tigecycline regimen was noninferior to the imipenem/cilastatin regimen for the c-mITT but not the CE population; this difference appears to have been driven by results in VAP patients.