If i knew nothing about .Net then i would rather learn 2005, by the time i'm on your course the Beta 2 will be out , the go live license will be available and it will be only months until the full version, better that, than learning a technology that's 2 years old already. If i already knew .Net well and was just learning the IDE then maybe my need would be more immediate and i'd want to learn 2003 now, but i think still i woould want to be learning 2005. and maybe my priority would be different depending on whether i was learning it for use in ASP.net or Windows Forms, as ASP.NET's take up will be quicker.

2003. It's real, there's a ton of info about it available that is "correct" and not speculation. Things for 2005 will change from beta to beta and even from beta to release. If you learn 2003, then there will only be a handful of things that you would have to come up to speed on for the next version. I think using a beta to learn a new technology is a bad idea.

I teach .NET classes and I don't think VS.NET 2005 is mature yet. See my post (dotnetjunkies.com/.../22284.aspx" rel="nofollow">dotnetjunkies.com/.../22284.aspx">dotnetjunkies.com/.../22284.aspx" rel="nofollow">dotnetjunkies.com/.../22284.aspx) on features that will no longer be present in ASP.NET in the upcoming beta 2 release.
Students are usually more cofortable learning technologies that have been around for some time.
Of course, if the intended audience are experienced .NET devs VS 2005 would be the tool of choice.

VS 2k3.

The demos that MS gave during the .NET 1.0 betas were mostly things that should never be done in a production environment. Best Practices and real world code didn't come for a while after .NET was released. During the betas, MS pushes the gee whiz features, not the real features (that use require a line of code to be written) and those gee whiz examples are not what a class should learn.

I doubt you would be teaching much beyond MS marketing if you were to have a .NET 2005 class this early.