arno -

I always recommend to use randomize plus rnd()^rnd(). Besides creating
"unique" numbers I think that many people try to use Excel random numbers
for some exact scientific analysis. This analysis could be ignored simply
by the fact of using excel as a random number generator that does not meet
scientific expectations to random numbers. I hope my recommendation
"improves" the random numbers of excel. <

The worksheet function RAND and the VBA function RND produce uniformly
distributed random numbers. That is, a number between 0.0 and 0.1 is as
likely as a number between 0.9 and 1.0.

The product RAND*RAND or RND*RND is random, but it is not uniformly
distributed. Products between 0.0 and 0.1 are much more likely than products
between 0.9 and 1.0.

However, I would greatly appreciate if someone could shed some light on
how "good" excel random numbers are, for what analysis they can be used
for and where one should not use excel. <

For such information about RAND, see
http://support.microsoft.com/default...b;en-us;828795

- Mike
www.mikemiddleton.com