n-3 Fatty acids have clinical benefits. The primary aim of the present study was the assessment of infection in patients who underwent major high-risk elective gastrointestinal surgery receiving postoperatively fish oil (FO)-supplemented parenteral nutrition (PN), compared with those receiving a standard olive oil (OO) emulsion. The secondary aims were the assessment of anti-inflammatory response and evaluation of tolerance and safety of these emulsions. A prospective, randomised, double-blind study was performed in patients requiring at least 5 d of PN. An isoenergetic and isoproteic formula was administered: group A received OO alone, while group B received OO that was partially replaced with FO (16·6 %, w/w). End points were outcome measures (mortality, sepsis, infection, hospitalisation days and PN duration), inflammatory response (C-reactive protein (CRP), prealbumin and leucocytes) and safety (TAG and glucose metabolism, and liver and kidney function). Statistical analysis was done using Student's t test and Fisher's exact test (P < 0·05). Twenty-seven patients were evaluated, with thirteen patients receiving FO. In this group, a significantly lower incidence of infections was found (23·1 v. 78·6 %, P = 0·007). CRP, prealbumin and leucocytes were not significantly different between the groups. There were no differences in safety parameters. We conclude that high-risk surgical patients receiving FO-supplemented PN for 5 d present a lower incidence of infection. Emulsions were safe and well tolerated.