Thanks guys! Ron, I tried yours because it was shorter. So it looks like
you are using counta to first judge if there are inputs in both cells. If
yes, then it is doing A1<B1. What is this function exactly? It seems to be
working great now, but I've never seen that notation before. Is A1<B1 just
a way to write "different"?



"Ron Coderre" wrote:

If your data is in adjacent columns (eg Col_A and Col_B)
Try this in Col_C:
=--IF(COUNTA(A1:B1)=2,A1<B1,0)

If the columns are separated, perhaps this:
=IF(AND(A1<"",E1<""),--(A10<E1),0)

In either case...
Copy the formula down as far as you need.
You'd only need to sum the test column.

Alternatively, you might consider this kind of single-formula approach to
count unmatched, non-blank items:
=SUMPRODUCT((A1:A20<"")*(E1:E20<"")*(A1:A20<E1: E20))

Adjust range references to suit your situation.

Is that something you can work with?
***********
Regards,
Ron

XL2003, WinXP


"pm" wrote:

Hello,

I have a list of 12,000 rows separated into two columns. In each row, the
two columns either match or don't. So I've made a third column that is just
a simple "if" function in each row that puts a "1" in if the two columns
differ. I then use a "countif" at the bottom to tally the # of different
rows.

The problem I am running into is that some of the rows have one column with
an empty cell (no data). So for this row I would have data for one column,
but not both. Obviously according to my functions, excel will count these as
different. I would like these rows to not be counted as different. Does
anyone know how to exclude rows where one of the columns has no data from
being counted as "different"?

As always, many thanks in advance.