I agree with you in general. But lots of times, I'll know that there's a column
that has to be filled in (some sort of key).

And if I know the data well enough, I'll use that method.

But if I don't know the data (some generic routine), I'll use the lastcell
(after trying to reset the usedrange).

I don't like the currentregion for the same reason you don't like the
..end(xlup).row. If there's gaps in the columns/rows, you might not get what you
really want.

(I like Debra Dalgleish's approach to find that real last used cell:
http://www.contextures.com/xlfaqApp.html#Unused)

Bill Renaud wrote:

"Dave Peterson" wrote:
<<If your data were in A99:A100, you'd get 2.

True, but then I was assuming that the application was for a list, since
this is where this technique usually shows up. Most people don't leave the
first 98 rows of a worksheet totally blank. (Unless they are some of the
posters to this NG! Grin! I try to teach good design principles rather than
giving the instant answer!)

Most of the other suggestions that were using some variation of
.End(xlUp).Row have the problem that not all columns will have complete
data. Say I have a table 5 columns wide and 100 rows deep (starting in cell
$A$1, of course!). If $A$100 happens to be blank, then any method counting
the number of rows of data only in column $A will return only 99, when the
100th row actually has data (in columns $B:$E). I guess that is why I always
use UsedRange, or CurrentRegion, or some variation of those properties, when
working with a list.
--
Regards,
Bill

--

Dave Peterson