Biff, I thought you had it. It seems that when the mod=2 is true, it
includes all the cells in the average b14,c14,d14,...,q14. The average does
not skip the cells where the mod=2 is false. Also the ideal would be to
include the cells with 0 or a number and exclude the blank cells. I could
proabably work that part out.

"Biff" wrote:

Try this:

Array entered using the key combo of CTRL,SHIFT,ENTER:

=AVERAGE(IF(MOD(COLUMN(B14:Q14),5)=2,B14:Q14))

To exclude cells that contain 0:

=AVERAGE(IF((MOD(COLUMN(B14:Q14),5)=2)*(B14:Q14<0 ),B14:Q14))

Biff

"jack" wrote in message
...
Thanks Biff, after I hit post, I realized that I should have given more
information. That was an example. I use most of the 256 columns
frequently.
It is monthly projection data. The Heading option actually works better
because of the unfixed columns (months can be added at the end of the
sheet
without having to change any parameters). An additional problem is that
for
an average of variable columns due to some tasks starting on different
months, a blank in the column would not be included in the COUNT and a
zero
would be included in the COUNT. ie. average = (sum of data) / COUNT

"Biff" wrote:

Hi!

If you only have 4 cells involved, what's wrong with:

=AVERAGE(B14,G14,L14,Q14)

Biff

"jack" wrote in message
...
I use multiple columns per month and would like to do an average (and
other
functions) on past months to forcast a rate for future months. Since
the
columns that I want to average are not adjacent, I cannot find a way to
do
it. Is there a function to identify the columns similar to the sumif
function, where I can identify the column based on a heading, and then
average (or stddev, variance) them. Another approach would be to index
into
or offset or indirect into the table selecting every fifth column ie.
average(b14:q14,5) average the values at b14,g14,L14,q14