Background: Intravenous anaesthetics and sedatives can influence polymorphonuclear tell (PMN) functions. Some of the drugs for sedation and anaesthesia have been alternatively dissolved in lipid solutions containing medium (MCT) and/or long chain (LCT) triglycerides. The in vitro effects of two different diazepam (benzyl-alcohol, LCT/MCT), etomidate (propyleneglycol, LCT/MCT), and propofol (LCT, LCT/MCT) preparations on respiratory burst (RB) and phagocytosis of human PMNs were studied. Methods: Diazepam (2, 20 mug ml(-1)), etomidate (0.5, 5 mug ml(-1)), and propofol (6, 60 mug ml(-1)) were investigated in clinical and 10-fold concentrations with flow cytometric assays. The RE was measured with the fluorescent dye rhodamine after induction with Escherichia coli or formyl-methionyl-leucylphenylalanine (FMLP) following priming with tumour necrosis factor alpha (TNF-alpha). Phagocytosis of PMNs was carried out in whole blood after incubation with fluorescein-labelled E. coli. Results: LCT-propofol at 60 mug ml(-1) reduced the percentage of PMNs with RE activity after induction with E. coli (52.8+/-20.4) and TNF-alpha /FMLP (10.8+/-5.1)) as well as the percentage of phagocytosing PMNs (48.9+/-19.5) in contrast to LCT/MCT-propofol, which augmented all parameters (85.4+/-10.1, 50.3+/-12.7 66.5+/-12.53. Also the higher concentrations of LCT/MCT-diluted etomidate and diazepam increased the percentage of RE positive PMNs compared to the alternative compositions. The percentage of phagocytosing PMNs was less reduced with 20 mug ml(-1) LCT/MCT-diazepam (85.2+/-6.9) than with the same concentration of benzyl-alcohol diluted diazepam (68.8+/-122) compared to the control. Conclusion: The in vitro effects of diazepam, etomidate, and propofol are dependent on the solvent applied. The tested LCT/MCT preparations reduce the inhibitory effects on the bacterial killing capacity of PMNs found after incubation with propyleneglycol, benzyl-alcohol, or LCT preparations, respectively.