(1) correct
(2) correct - actually, I change the 'Underlay' to be the fit and use Dataset#N to overlay the time course data in the plot window.
(3) restores the problem. Once there is a "bad" fit data set, it stays "bad", unless I restart afni.

A couple of things

1- I'm not sure, but it seems that sometimes the fit set is just scaled incorrectly, not completely incorrect. I create the fits and write an afni BRIK.gz (and HEAD) using IDL. The underlay (the time course data) are also created by me and written in IDL. They are both stored as float (not scaled integers). I (try to) create the unique id codes - could this be an issue?

2- I have AFNI_RESCAN_METHOD = REPLACE

3- These symptoms also happen using Dataset#2 instead of Dataset#N

Version AFNI_2007_05_29_1644
I compiled, using suncc_64
SunOS 5.9