Can anyone tell me what's going on here? I've got a client that's sending an email from a windows system that is a set of data in text format that's sometimes has an empty tab/space character at the end to represent a null set. Every line like this has an "=20" at the end and the line breaks are getting all messed up...so it's difficult to even replace all the =20's with the proper tab character or whatever.

Any idea how to prevent the email from getting broken up like this? I thought at first it was a windows to linux thing, but dos2unix does nothing to my file.

This is with Sendmail 8.14.0 on a Redhat 4.0 box by the way.

What you are seeing is Quoted-printable encoding. (And the fact that you see it means that either the sending or the receiving mail client is broken.)
But you should be able to clean it up with a simple search/replace. Maybe the mangled line breaks are soft breaks? (See the linked article.)

Thanks, that gets me a step closer. I found an online convert for qp encoded text, and that seems to convert it. But ideally, I'd like to stop it from happening in the first place. Any idea on that?

I'm pretty sure the email is coming from Microsoft Outlook on the windows system. I can recreate it from my laptop using Outlook and it happens every time. Is there a way to tell Outlook to stop using qp encoding? on the linux end, I'm just viewing it using the mail command.

Well, you might get Outlook to not send quoted-printable, but that's not necessarily a good thing. QP is there for a reason as it will protect your mail from possibly being mangled by any server it passes through.

Better would be to find out why your client doesn't decode or, if it doesn't do so by design, run the mails through a decoder yourself.