How can I use covariance in random number generation
of a normal distribution?

For example, I would use VBA to implement Box-Mueller
using mean and std dev. I have implemented B-M in C.
Can someone provide an algorithm (VBA, C or pseudocode)
that incorporates covariance between two and among
3 or more random variables, each normally distributed?

Is that called multivariate normal random number
generation?

FYI, This is for Monte Carlo simulation, if that matters.
And I would be content with (even prefer) an algorithm
that depends on a uniform RNG.

Also, can I accomplish this without using VBA?

For example, some people have suggested using
NORMDIST(RAND(),mean,sd) as an RNG with normal
distribution based on mean and std dev. I don't know
how good that is and how that compares quality-wise to
Box-Mueller, even assuming that RAND() is a good
uniform RNG. (I believe some people question that
assumption.)

Frankly, I don't know how good Box-Mueller is quality-wise
either, assuming a good uniform RNG to start with. I
believe some people quibble over B-M as well. But I
prefer to keep the discussion simple for now. I am
interested in fundamental algorithms, not quibbles about
perfection.