Hi Jim,

I see that you got your aanswers with Cetin's help.

>
>>Expanding on my previous observation, days applied first, month second and years last. I'm not sure is it better than previous one, though. :)

dob = GOMONTH(GOMONTH(dt-dd, - mm), -yy*12)

>
>Me neither ;-) I decided to test against some random dates and ages as below. It is surprising how seldom that the methods agree, and the variance is far wilder than I expected. Do you see anything wrong with what I am doing?
<snip>