1 INTRODUCTION Animal models of human diseases are a widely used research tool to understand the progress of diseases and to evaluate potential therapies and new drugs (Cherry, S. R. & Gambhir, S. S. 2001) (Luanne, L. P. et al. 2007). The use of animal models has recent- ly gained increased interest due to the availability of in vivo small animal imaging and to the rapid growth in genetics and molecular biology (Phelps, M. E. 2000) (Cherry, S. R. & Gambhir, S. S. 2001). Positron Emission Tomography (PET) is a non- invasive nuclear medicine technique allowing the measurement of the spatial and temporal distribution of radiotracers which map physiological and meta- bolic functions of the body (Wernick, M. N. & Aarsvold, J. N. 2004). Since biochemical changes precede morphologic changes, PET has the potential to provide diagnostic information earlier than, for example, X-Ray Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) (Gambhir, S. S. 2002). The use of PET in small animals allows the use of subjects as their own control, reducing the intera- nimal variability. This allows performing longitu- dinal studies on the same animal and improves the accuracy of biological models. Small animal PET is becoming very useful in several fields of molecular imaging such as the development of techniques to measure endogenous and reporter gene expression in vivo. Moreover, the ability of small animal PET to map the in vivo biodistribution of new pharmaceuti- cals opens up new possibilities for research using small animal models of human diseases (Cherry, S. R. & Gambhir, S. S. 2001) (Cherry, S. 2004). How- ever, small animal PET still suffers from several li- mitations. In fact, the amounts of injected dose needed, limited scanner sensitivity, image resolution and image quantification issues (Chatziioannou, A. F. 2002) (Tai, Y.-C. & Laforest, R. 2005), could clearly benefit from additional research. The simulation of small animal imaging using Monte Carlo allow modeling imaging systems, de- veloping and assessing of tomographic image recon- struction algorithms and evaluating correction me- thods for improved image quantification (Andreo, P. 1991) (Zaidi, H. 1999). In this context, Monte Carlo simulations are becoming an essential tool for assist- ing this research and some specific Monte Carlo si- mulation packages have been evaluated for nuclear medicine applications (Buvat, I. & Castiglioni, I. 2002) (Buvat, I. et al. 2005). Recently, the Geant4 Application for Tomographic Emission (GATE) platform has been developed (Jan, S. et al. 2004) and validated for the simulation of the microPET ® FO- CUS 220 system (Jan, S. et al. 2005). Small Animal Simulation Studies using the microPET ® FOCUS system and the GATE platform S. Branco & P. Almeida Universidade de Lisboa, Faculdade de Ciências, Instituto de Biofísica e Engenharia Biomédica, Portugal S. Jan CEA/DSV/I 2 BM, Service Hospitalier Frédéric Joliot, France ABSTRACT: GATE, a Monte Carlo simulation platform dedicated to nuclear medicine, has been validated for modeling μPET systems like the microPET ® FOCUS 220 under realistic imaging conditions. In this paper, the use of the microPET ® FOCUS 220 simulation model with GATE is used along with realistic mouse phan- toms (including respiratory motion) and radioactive distribution maps, obtained from real exams, to produce simulated PET data of a mouse. We present results from simulated realistic whole body studies of the mouse with [ 18 F]fluoride and 2-Deoxy-[ 18 F]fluoro-D-glucose (FDG) and compare them to real data. The simulations were performed using voxelized mouse phantoms, obtained from real PET examinations and from digital mouse anatomical models. Respiratory motion is introduced in the simulation process. The qualitative and quantitative results from simulated data reproduce real values within 23%. This discrepancy can be related with specificities of simulated and real data, in particular due to attenuation and scattering corrections. Keywords: Monte Carlo method; GATE; microPET ® FOCUS; small animal; PET imaging