There's no doubt that by now you've noticed that we're big fans of R around here. It's completely free, has superior graphing capabilities, and with all the extension packages available there isn't much it can't do. One of the problems with R especially to new users is that it isn't obvious how to find help when you need it, so I'll try to provide a few resources here.
If you're completely new to R but you know some statistics, I've mentioned this book once before - Introductory Statistics with R by Peter Daalgard ($48, Amazon). I was able to get a good feel for using R after only a few hours with this book. For some free help to get started, check out my previous post covering Theresa Scott's R course, which provides links to her website and her course materials. One reader commented that Theresa's tutorials were better than any R book he'd ever read.
If you've used R before but need a quick reference on the arguments for a particular command, just type a "?" followed by the name of the command: ?rnorm.
What's a little more difficult is figuring out how to do something in R where you don't know the commands or packages that will do it. One option is a Google search, but "R" is not a very helpful search term, and you'll often find irrelevant results. To deal with that, some R users have developed a custom Google search at Rseek.org, which will perform a Google search for your terms but only return results from sites known to contain R help and documentation.
Finally, check out the previously mentioned R clinic at Vanderbilt. Ever week Theresa Scott in biostatistics holds a clinic and allows R users to bring in specific questions and/or their own data for help.