e20506 Background: We have previously shown (dos Santos JNCI 2012) that NK1 receptor antagonists (NK1RA), such as aprepitant and casopitant, improve chemotherapy-induced nausea and vomiting (CINV) at the cost of increasing severe infection risk. Herein, we update our results. Methods: We searched MEDLINE, EMBASE, CENTRAL and meeting proceedings to identify randomized controlled trials (RCTs) comparing standard antiemetic therapy (dexamethasone + 5-HT3R inhibitors) versus NK1RA plus standard therapy for CINV prevention. Complete response (CR) was defined as absence of emesis and rescue therapy. The endpoints were CR post-chemotherapy in the overall phase (first 120 hours), CR in the acute phase (first 24 hours), and the delayed phase (24–120 hours), nausea, and toxicity. Subgroup analyses evaluated the type of NK1RA, the emetogenic potential of the chemotherapy, and prolonged use of 5-HT3R inhibitor. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using a random-effects model. Results: A total of 20 trials (9,935 patients) were included. NK1RA increased CR rate and nausea in all phases and subgroups. The CR in overall phase improved from 56% to 72% (OR = 0.52, 95% CI = 0.46 to 0.59, P < .00001, NNT=6). The side effects of NK1RA followed our previous report: the severe infection rate increased from 2.4% to 5.6% in the NK1RA group (four trials, 1656 patients; OR = 2.87; 95% CI = 1.58 to 5.19, P = .0005, NNH=31). Conclusions: NK1RA consistently improve CINV control in the acute, delayed, and overall phases, either for moderately or highly emetogenic chemotherapy regimens. Their use might be associated with increased infection rate, and a patient-level meta-analysis is warranted to clarify this safety issue.