Axioms are the things that are taken as basic, unchallenged assumptions. Depending on what area of mathematics you are working within, these may change. For instance, all of the elementary results of arithmetic are usually taken as unstated axioms in higher branches of mathematics (so you don't prove $1+1=2$ in a calculus course, but might in a certain other courses).
Theorems are conclusions that can be drawn from a set of axioms by using the rules of logic. Not every such conclusion is bestowed with the title of theorem though: usually the conclusion has to be meritorious, and often the verification is non-trivial.
I've always thought of postulates as being slightly less fundamental than axioms, but nevertheless similar in their nature as assumptions from which other statements are to be proven. To give a natural language analogy: I take as an axiom is that I exist, and I postulate that my senses aren't being manipulated as part of a perfidious plot.