Septicemia causes multiple and often not very well understood metabolic alterations. In this sense it is controversial whether or not carnitine is decreased, which may have several implications. Our objective is to measure the plasma carnitine levels in septicemic patients, and to find out whether this is modified if there is a multi-organ dysfunction syndrome (MODS), or if it is dependent on the lipid source. For this we carried out a prospective study in septicemic patients with MODS. These were given exclusively parenteral nutrition (PN) without any carnitine supplementation. The PN of 16 patients contained long chain triglycerides (LCT's), while that of another 12 contained a 1:1 mixture of long and medium chain triglycerides (LCT's and MCT's). We measured the plasma carnitine level at the baseline, after 5 days and after 10 days, using an enzymatic method that measures the total carnitine level (free and esterified). The normal values lie between 35 and 70 mumol/l. We included 28 septicemic patients whose ages were 53.41 +/- 16.51 years, and whose APACHE II on admission was 17 +/- 4. The carnitine levels were: baseline: 60.4 +/- 23.7; 5th day 57.7 +/- 22.9; 10th day 55.7 +/- 21.2 (p = n.s.). Of these patients, 10 had an MODS of septic origins with their baseline levels being: 65.3 +/- 30.9; 5th day 60.3 +/- 23.2; 10th day 61.5 +/- 15.5; while the levels of the 18 septicemic patients without MODS, the baseline levels were 61.9 +/- 13.8; 5th day 58.6 +/- 19.1; 10th day 56.6 +/- 19.3 (p = n.s.). In the patients who were given LCT's the baseline carnitine level was 60.7 +/- 23.1, 5th day 60.1 +/- 23.8; 10th day 58.6 +/- 12.8; while those patients who were given LCT/MCT showed baseline levels of 64.3 +/- 19.5; 5th day 58.6 +/- 19.1; 10th day 57.8 +/- 10.7 (p = n.s.). In our septicemic patients the serum carnitine levels we found were within the normal range, and these remained unchanged during the ten days in those patients with MODS or with the lipid mixture used.