Speaking as a Master’s degree holder, i’m biased but I’d say the advantages are primarily:

1) Cross-college connections and alumni networking, especially if you go to a “brand-name” school. Sorry to offend or seem elitist but it’s true.

2) The opportunity to do creative, exploratory projects and re-kindle the imaginative spirit that the working world may have killed off (Like Jack I went straight thru from Undergrad to Grad, for various reasons, but I remember my CMU adviser saying he liked folks who returned to school after spending a few years in the “real world” b/c they were sufficiently angry and jaded and primed to crank out amazing stuff—i’m simplifying a bit ;-)

3) The opportunity to get deep into thinking, reflecting, and diving into the theoretical and intellectual issues that enrich the practice, but we often don’t have time for when we got a 12pm deadline for a client and then a proposal due at 5pm. Spending the year or two doing that deep dive (if you really enjoy it—alot of folks admittedly don’t) may help cultivate a valuable habit that will make returning to the real world a bit more tolerable and satisfying. The intellectual fodder you gain does provide valuable perspective. At least that’s what I tell myself when engineers are clammoring for specs yesterday and I have to design for the PM’s delusional use cases :-)

4) And if you’ve been fumbling around learning it as you go along, grad school offers the chance to learn methods/approaches in a more organized guided fashion (presuming the curriculum is sound and robust!) to push yourself further…and perhaps discover something about yourself you didn’t know!

Also, in terms of career growth, AIGA and IDSA usually publish periodic studies of salary increases, etc. More and more I see job descriptions (like posted on ixda) that require or recommend Master’s…

That all said, in the end it’s a personal choice and has to be measured against your passion and what you really want to get out of the degree. And if it’s right at your stage of life, career, etc.