To determine the influence of peripheral parenteral nutrition (PPN) on the incidence of phlebitis and the effect of simultaneous infusion of lipid emulsion, 68 surgical patients were assigned randomly to 3 nutrition groups: (1) amino-acid 4% solution and dextrose 8% (AA + DX) + Intralipid((R)) (IL) 10% in the ratio 1.5:1 (2) AA + DX and IL 20% in the ratio 3:1 (3) AA + DX alone. Group 4 (control group) consisted of 25 patients receiving maintenance solutions and iso-osmotic drugs. No filters were used. The cannulae were studied for 72 hours and the cannula sites were evaluated daily for phlebitis. Mean age of patients was 52.3 +/- 17.7 years, with a similar sex ratio in the 4 groups. One hundred and twenty-five cannulae were studied. At 72 h, the following cumulated phlebitis rates were recorded: 22% in Group 1 (712 mOsm/L), 48% in Group 2 (802.5 mOsm/L), 44% in Group 3 (920 mOsm/L) and 26% in Group 4 (260-314 mOsm/L). The phlebitis incidence was similar in groups 1 and 4 (control) throughout the study. Both Groups 2 and 3 had higher phlebitis rates on days 2 and 3 (II: p < 0.05, III: p < 0.01). It is concluded that PPN with simultaneous IL 10% in the ratio 1.5:1 is no more phlebogenic than the usual maintenance solutions during the first 72 h. This is probably due to the lower osmolality and lower DX and K concentrations of the solution with IL 10%. Intralipid((R)) 20% does not have the same protective venous effect. The lower complication rate, the limited need for nursing, and lower costs compared to central parenteral nutrition, make PPN a safe nutrition technique for up to 10 days in patients needing no volume restriction.