Hey Pete,

Well now that you have the source file, you can understand the kind of
situation i am dealing with. The order matters because of the other
data attached to each value.

As for the one-to-three way cancelling, im not sure exactly what i may
have said before, but the definitive version is that one value will
cancel with its opposite once. There may be multiples of that value
in a + or - fashion, but if there are an even number of each sign,
all of the values will probably cancel.

As for the macro running directly in the source file, it would be
useful. I could just sort the numbers how i saw fit, and then run the
macro right there. Unfortunatley, dates and value sizes arent the
only things i need to worry about when developing a solution to
cancelling out values without making any mistakes. Its hard to cover
all of the bases because there are so many variables...date of
posting, current period, journal status (N, B, or R), etc.

If you wanted to alter the macro to make it run in the source file
(which resembles what i sent you), i would welcome that change. It
would be nice if you could just add comments to the code so i knew how
to change which columns it looked at, etc...

I will revert the tenths to the hundreths place in the macro's search,
that was my mistake for changing it, it is better off being more
precise.

As for my earlier problem of the wrong pairs being created (unique N's
being highlighted instead of the B and R entires that should cancel),
it is a pretty easy fix manually, all i do is sort out the uncancelled
values, and sort by Journal type to weed out all the R's which i know
should cancel. Then i find their corresponding N value which was
highlighted mistakenly, and just reverse the highlighting so the
correct one is left unhighlighted (N).

I believe a combination of resorting the data, possibly in the way in
which you suggested, along with comparing numbers to the hundreths
place, and possible having the macro pay attention to journal entry
type (N, B, R) will solve this problem.

Let me know once you have looked at the dataset if you had any ideas
as to how to cater to this particular issue.

Thanks again for all of your help Pete.

-Pogster