Thanks for the reply, Karl. That was very helpful. I re-looked at our data to check on the question of volume and it raised new questions to my mind. I think, perhaps, I'm missing something rather basic... The volumes in our PET and fMRI data are different but they seem to vary with the smoothing I applied. For instance, I smoothed both data sets at 8, 12, and 16mm. The search volumes (in voxels) are displayed below: PET fMRI 8mm 187833 208800 12mm 213866 213245 16mm 237025 217173 (all voxels were isotropic 2mm^3). To me it seems like the differences in volume were small which would make pretty good sense -- in both our PET and fMRI scanners we tried to acquire whole brain images with approx ~15cm FOV. What's odd, though, is that the calculated FWHM smoothness of our data (in mm) are: PET fMRI 8mm 8.0 9.4 10.8 15.8 14.9 10.7 12mm 11.0 13.0 14.7 22.7 20.9 13.8 16mm 14.5 17.1 19.1 23.5 21.9 15.5 Am I correct in reading these values as indicating that the fMRI data was smoother than the PET data? - Joe %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%