In this paper measures and functions on GL(n) are called bi-invariant
if they are invariant under left and right multiplication of their arg
uments. If v is any bi-invariant Borel measure on GL(n), then there ex
ists a unique Borel measure v on D-+greater than or equal to(n), the
set of all diagonal matrices of rank n with positive non-increasing di
agonal entries, such that [GRAPHICS] holds for each v-integrable bi-in
variant function f:GL(n) --> R. An explicit formula for v will be der
ived in case v equals the Lebesgue measure on GL(n) and the above inte
gral formula will be applied to concrete integration problems. In part
icular, if v is a probability measure, then v can be interpreted as t
he distribution of the singular value vector. This fact will be used t
o derive a stochastic version of a theorem from perturbation theory co
ncerning the numerical computation of the polar decomposition.