Rick,
Yes, I just left that out of my last post. But, I had no trouble breaking up the data into runs, running it through 3dMaskAve and 3dDetrend, and then pasting it back together with 3dTcat.

As far as using 3dDeconvolve, I actually tried something similar to what you suggest here after your first post. I broke the data up into 2 halves and ran 3dDeconvolve on each half; then I put the two halves back together and used 3dMaskAve. The result didn't seem to have removed any components.

However, just now I tried what you suggested (basically, just flipping the order of the 3dDeconvolve and 3dMaskAve commands). Seems to have worked very well; and, as promised, the output has no headers to deal with. The only issue was the -mask option. My 3D mask is no longer in the same dimensions as the 1D data. Since it isn't useful anymore (given that the data have already been masked), I just removed it.

Thanks again for all your help today,
Adam