Are you agnostic or atheist? Perhaps you are both.
Politically and socially, the two words can represent a desire to take or not take a particular position on the issue of God. You can self-identify as an agnostic to indicate you're just not interested in discussing the issue. If you self-identify as an atheist, you're declaring at least that you do have a definite position.
There's nothing at all wrong with self-identifying politically and socially as an agnostic. If you don't want to discuss the issue, if you don't want to take any particular position (at least in public), you're free to do so. If you don't believe that a god exists, I would certainly encourage you to publicly self-identify as an atheist, but I won't fault you if you choose otherwise.
If you're interested in the philosophical implications of your choice, however, you might find the rest of this post informative.
Philosophically, the words agnostic and atheist (and their derivatives) represent two very different concepts. Agnosticism is about knowledge; atheism is about existence. In philosophical jargon, agnosticism is an epistemic term; atheism is an ontological term.
If you're a philosophical agnostic, you're saying that you don't know whether or not a god exists. If you're a philosophical atheist, you're saying at least that you don't believe a god exists (weak atheist) or that you believe no god exists.
Right off the bat, if you don't know whether or not a god exists, you are therefore either a fideistic theist (you believe in god on faith, not knowledge) or you are at least a weak atheist: if you don't know, and you don't believe on faith, then you don't believe. Many religious people consider lack of belief to be as atheistic as you need to be to be at least excluded from the religion and at worst susceptible to eternal damnation.
Agnosticism doesn't just apply to god; in general uses, there are two distinct meanings: 1) possible to know but not actually known; and 2) impossible (or deeply impractical) to know.
An example of agnosticism in the first sense would be about the present or historical existence of life on Mars. It is (or seems) possible to know; we just have not yet looked particularly thoroughly. The obvious conclusion from this sort of agnosticism is &mdash if you're interested in the question — to gather more evidence, to look more thoroughly.
In the second sense, it's impossible to know that there aren't really invisible elves pushing us towards the center of the earth, instead of the prosaic gravitational field (or the space-time warping of General Relativity). It's also impossible know how many Apaches are hiding in this room (if we knew about any, they wouldn't be hiding, they'd be failing to hide). It's possible in the most general principle to know that there's no extraterrestrial life in the observable universe (since the observable universe is finite) but it's deeply impractical (at least by present technology) to actually look.
There are more subtle forms of unknowability. Godel's second incompleteness theorem proves that we can't know that the inverse of a theorem in arithmetic is not derivable. We know that 2+2=4, and we know we can know it (and we can know we know, etc.), but we cannot know that it's impossible to derive the contrary.
The status of statements whose truth or falsity is impossible in principle to determine by any means is a matter of no small philosophical controversy. One the one hand, we do consider truth to be independent of our knowledge: there are true statements (or so we intuitively believe) which are true even if we don't know they're true, or even if we think we know they're false: we can be ignorant of or even mistaken about the truth. But this construction would seem to apply most strongly to knowledge that's deeply impractical, or statements we have evaluated in a biased, lazy, restricted or otherwise deficient manner.
We cannot, however, blithely extend our beliefs about impractical or deficient knowledge to impossible knowledge. The proof is simple. We can recast our beliefs about impractical or deficient knowledge exclusively in terms of justification. There are statements we cannot not presently justify one way or the other that we could justify, at least in principle. (This principle applies, mutatis mutandis to mistaken justification, as well as finding the "right" answer with the "wrong" justification.) We cannot, however, cast our beliefs about impossible knowledge exclusively in terms of justification. It is a blatant contradiction to assert that there are statements we cannot in principle justify that can in principle justify.
Of course, that's not to say that statements impossible to know aren't truth-apt; it's to say only that the extension of our beliefs about impractical knowledge require an enthymeme to extend to impossible knowledge. You can assert without contradiction both that statements impractical to know are truth-apt and statements impossible to know categorically are not truth-apt. (You deny the enthymeme that connects the two cases.) And you can do so without giving up your ordinary notions about prosaic objective reality. (It does appear to be the case, however, that asserting the truth-aptness of statements impossible to know underlies a considerable portion of — and perhaps even the entirety of — theological and philosophical metaphysical bullshit.)
(There are still some subtleties: What about propositions that can be justified, but require an infinitely long justification? How about propositions (such as an individual Godel statement) not provable in a language, but provable in a meta-language? These are side issues though; the sort of god that requires advanced transfinite mathematics or meta-mathematics just to define does not seem the sort of god the vast majority of religious believers are talking about; not the sort of god obsessed with the correct enjoyment of our naughty bits.)
One problem with the meaning of "agnostic" that crops up in ordinary conversation (especially between theists and atheists) is an equivocation between knowledge and certainty. In ordinary conversation, we say we "know" things about which we are not certain, especially if we grant — as most atheists do grant — that professional scientists are gaining actual knowledge. It doesn't appear that we know with absolute certainty anything that passes informally or intuitively for knowledge, at least not with our finite minds. If we construct "agnostic" to mean "not absolutely certain" then we must admit agnosticism about everything, rendering the term vacuous.
(Another problem is the equivocation between not known positivistically but known on general principles. If you claim that you dropped something and it didn't fall, even though I didn't actually see it (not) happen, I know that your statement is false precisely as well as I know the law of gravity is true. Although Russell probably intended his teapot example as a metaphor, he missed an epistemological trick: his example can be known to be false on general principles without positivistic knowledge.)
The truth-aptness of the impossible to know and the equivocation between known and known-with-certainty forms a fallacious argument trotted out in various guises by theists: You cannot "know" (i.e. you can't be certain, or "god exists" is unknowable) god exists that god does not exist. Therefore your belief that god does not exists is false. Therefore god exists. Stated clearly the underlying fallacy is obvious (you don't "know" that there's no elephant in your living room, therefore your belief that there's no elephant in your living room is false, therefore there's an elephant in your living room); rhetorical skill can, however, obscure the fallacy.
(The basis of this fallacy is often invoked with confident assertion that "you can't prove a negative." This assertion is trivially false: Euclid, for example, proved that there is no largest prime number. Many negatives can be proven, and, depending on which standard of proof you apply, there are many positives that cannot be proven.)
It all boils down to three cases:
1: It is impossible to know "a god exists" or "no god exists"
In this case, it is entirely legitimate to conclude that both statements are entirely meaningless, or at least meaningless as to truth (i.e. saying "God exists" is like saying, "Yay!") If you think "god exists" is meaningless, you definitely do not believe that god exists, and you're an atheist.
2: It is possible but impractical to know whether "a god exists" or "no god exists".
Who cares? Presumably a god who wants itself to be known will not hide; contrawise a god that makes itself difficult to know does not want to be known. It's definitely the case that the characters in most theistic scriptures did not hide at all.
3: It is practical to know whether "a god exists" or "no god exists", but not with certainty.
In this case, we know at least as well as we know anything that no god exists. If you're going to refuse strong atheism on this case, then you should refuse to acknowledge that no elephant exists in your living room. (You might, for example, be hallucinating or subconsciously denying your perceptions whenever you look. In which case I can get you a great deal on an elephant shit shovel; email me.)
In the philosophical sense, if you are agnostic about god, you are (if not a theist) at least a weak atheist. If you believe that the existence of god can be known practically but without certainty, you are as much a strong atheist about god as you are about elephants in your living room.