I think I understand a theorem if I can reconstruct the hypotheses remembering only the conclusion. Likewise, I think I understand a theory if the axioms all seem reasonable and obvious.
However, you can't really get this on a first reading. I usually allow myself to proceed if the proof of the theorem catches my interest enough to read the details, or seems too trivial to bother with. If I don't think I could do it but don't care to learn, then I obviously don't appreciate enough of what I have been reading to seriously try reading more.
Perhaps the ultimate test is if you can use said theorem for something.