Combination therapy for infections with Gram-negative bacteria (GNB) has been the subject of debate over years, with studies leading to different and at times conflicting outcomes. Proponents of combination therapy propose several arguments: efficacy with expanded broad-spectrum coverage provided by two antimicrobial agents with different spectra of activity; prevention of the emergence of resistance; in vitro synergy between two antibiotics, and improved mortality.Most of the reviewed published data included retrospective studies, and no randomized controlled trials were found. There seems to be a consensus that in sensitive GNB, combination therapy has no added benefit in improving mortality. Hence, the widely used practice is to de-escalate the coverage as soon as the cultures grow a sensitive organism. However, in the era of multidrug resistant (MDR) bacteria and the absence of new antimicrobials in the pipeline, combination therapy should be readdressed.Most of the available data derive from studies involving Pseudomonas aeruginosa , carbapenem-resistant Enterobacteriaceae (CRE), and Acinetobacter baumannii . The results of clinical studies on combination therapy in the treatment of P. aeruginosa are conflicting and non-conclusive. One of the largest recent retrospective cohort studies found no difference in 30-day mortality between combination and monotherapy. However, the isolation of an MDR organismwas not specifically examined. In another study, in vitro activity against MDR P. aeruginosa demonstrated the effectiveness of antimicrobial agents when used in combination, and that despite documented resistance, bacterial killing may still be achieved with some combinations. As for CRE, several retrospective and prospective observational studies showed improvement inmortality with combination therapy. In vitro studies showed that the combination of colistin/tigecycline, rifampin/polymyxin, tigecycline/meropenem, and imipenem/colistin had synergistic activity and fosfomycin/carbapenems prevented the emergence of resistance. Recent studies suggested that combination therapy of 2 or more agents was associated with less mortality. Extended infusion of meropenem in combination with other drugs seemed to have the best outcome when the MIC is <4 ug/ml and a combination of colistin, meropenem and tigecycline had the best outcome regardless of the MIC to meropenem. In addition, a recent study suggested that a combination of 2 carbapenems – ertapenem and doripenem – was effective in treating infections with CRE. Finally, for MDR A. baumannii , the clinical information regarding the use of combination therapy is scarce and is limited for themost part to in vitro studies. In view of the extremely limited armamentarium, few antimicrobials remain active against MDR A. baumannii . Polymyxins and tigecycline are such drugs, and both are recommended for use in combination with other drugs. In summary, the available active agents againstMDRGNB aremainly the same: polymyxin, tigecycline, systemic fosfomycin, carbapenems and rifampin. However, their limited PD/PK makes them less potent for the treatment of serious infections. Hence their use should be avoided as monotherapy in the treatment of severe infections. Symposium 17. How to use PK–PD for our best practice