Leukocyte activation is a property of systemic infection. In a previous animal experiment we could demon- strate, that granulocytes, which were already activated by the immune system, were more useful in the evaluation of scin- tigraphic imaging infection than non activated donor granulocytes. The purpose of the present study was to investigate the in vitro activation of isolated polymorphonuclear (PMN) donor leukocytes in the presence of various biological active modulators in rabbits with E. coli infection. Methods: In vitro, incubation of isolated leukocytes of non infected donors was performed with different immune stimu- lating modulators such as granulocyte-colony stimulating factor (G-CSF), proinflammatory cytokines (IL-8, IL-1 ) and bacterial products (fMLP) at 37 degrees C for 2 hrs. Afterwards, the different radiolabeled granulocyte preparations were studied in rabbits with an E. coli infection in the left calf muscle. The soft tissue infections were scintigraphically visual- ized following injection of 18 MBq 99m Tc-HMPAO-purified-heterologous in vitro stimulated granulocytes of non infected donor rabbits. Non-stimulated 99m Tc-HMPAO-purified-heterologous granulocytes served as a control. Gamma camera im- ages were acquired at 2 min, 1, 2 and 4 hrs p.i. After the last image the rabbits were sacrificed and the uptake of the radio- label in the dissected tissues was determined. Results: The 99m Tc-HMPAO-heterologous granulocytes incubated with G-CSF faintly visualized the infectious focus in the calf muscle at 2 hr p.i. and a slightly better delineation of the infection was noticed at 4 hr p.i. With heterologous granulocytes incubated with proinflammatory cytokines and fMLP a delineation of the infected calf muscle was not pos- sible. The absolute uptake in the infected calf muscle was not significantly different between G-CSF (0.26 ± 0.06% ID), proinflammatory cytokines (0.23 ± 0.06% ID), fMLP (0.22 ± 0.02% ID) and the controls (0.17 ± 0.02% ID). The ratio of the infection to the non infected contralateral muscle was slightly higher for 99m Tc-HMPAO-heterologous granulocytes incubated with G-CSF (2.63 ± 0.03) as compared with the proinflammatory cytokines (1.3 ± 0.01), fMLP (1.4 ± 0.08) and the control (1.4 ± 0.04), respectively (all statistical differences were not significant with p=0.1-0.32). Conclusions: Our results confirm a direct, but probably weak stimulating effect of only G-CSF on the PMN for imaging infection. In addition, in-vivo heterologous granulocytes harvested from infected animals in a previous study showed bet- ter results as compared to the present data, suggesting the need of intrinsic cell activation for specific granulocyte migra- tion, which cannot be mimicked by other stimuli.