how to model a dependent variable that is a share with many 0 's AND many 1's?

Hi there,
My dependent variable is the share of credit card to total credit, it ranges from 0 to 1 and it is concentrated in both 0 and 1 extreme values.
What model do you suggest to handle it?
Many thanks in advance!

I would try beta regression, Beta Regression in R (r-project.org)

thank you nirgrahamuk, but Beta Regression doesn´t take into account the extreme values: 0 and 1 that are precisely the particularity of my dependen variable
any other model would be suitable??

Maybe then zero one inflated beta regression , I see theres a zoib package for that.