The Charmides begins with Socrates arriving back in Athens after years of service in the army and a recent escape from a brutal battle. He heads for a palaestra to find his old friends, who ask him about the battle. He asks them, in turn, about the state of philosophy in Athens, and whether there are any particularly wise or beautiful youths there at the moment. One such youth, of unsurpassed beauty and solid temperance, is said to be Charmides, who enters the temple almost immediately. Socrates is awestruck and aroused by Charmides's beauty, and manages to persuade Critias (Socrates's friend and Charmides's older cousin) to convince Charmides to come over by telling him that Socrates has a cure for Charmides's headaches. Socrates persuades Charmides that the mystical "cure" treats not just the body but the soul as well, making the soul temperate.
Charmides agrees to receive the charm, but Socrates first must make an "inquiry" to find out if Charmides is actually already temperate of soul (as everyone says he is). The inquiry proceeds with Charmides attempting to define temperance. He tries to define it as "quietness" and "modesty," but both are shown by Socrates not to match the definition of temperance (which, Charmides concedes, is always good): quietness is not good in some pursuits, and modesty is sometimes bad (when one is needy, for example). Charmides then uses a borrowed definition: temperance is "doing our own business." Socrates refutes this as well, on the grounds that craftsmen, for example, can be temperate even though they do, or make, things for other people.
At this point, it becomes clear that Critias is the source of the "doing our own business" definition, and he is brought in to defend it (Critias is the interlocutor from here until just before the end of the dialogue). Critias begins with a defense of his definition based on a nitpicky distinction between doing, making, and working, which Socrates proceeds to put in simpler terms: temperance, Critias is arguing, lies in doing good. Socrates expands this proposition to involve doing good both for others and for oneself, but quickly shows that people often do not know which of their actions will be beneficial in these ways. Thus, it seems possible to be temperate without self-knowledge.
Critias objects strongly to the suggestion that one might be temperate without knowing that temperance, and revises the whole argument, making "self-knowledge" the new definition of temperance (he introduces this new theme by quoting the oracle at Delphi). The two debaters decide that, if temperance is a kind of knowledge, then it must be a kind of science; Critias's initial suggestion is that temperance is "the science of a man's self" (though it remains unclear, as Socrates points out, what product or effect such a science is supposed to have). At this point, there is a brief but significant interruption in the argument, as Critias accuses Socrates of trying only to refute whatever he says rather than develop the argument. Socrates replies by noting that refutation has nothing to do with positive arguments attached to individuals; it is rather precisely the way the argument does proceed.
The assertion that temperance is a kind of self-knowing science is developed into the hypothesis that temperance must be a science of itself and of all the other sciences, a "science of science." Further, Socrates adds, in order to be truly self-knowing, such knowledge must also know what it does not know. This "science," of temperance, then, seems to be a very difficult notion to conceive of (and perhaps even a paradox)—Socrates is unsure if such a knowledge can exist at all, and he is further unsure if it would do any good at all. The two decide to drop the question of whether such a "knowledge of knowledge" can exist at all, and focus on whether, if it did exist, it would be of any use.
In seeking a use-value for this kind of reflexive knowledge, it rapidly becomes clear how difficult it is to conceive of any possible links between knowledge whose "subject matter" is purely knowledge and knowledge that has concrete subject matter (like medicine or architecture). Socrates suggests that perhaps he and Critias were aiming too high in defining temperance strictly in terms of pure-idealized self knowledge; how, after all, can pure self-knowledge rule over all knowledges of concrete people and things? Socrates proposes a different, less ambitious version of the same idea: knowledge of knowledge helps frame or guide more concrete learning.
This revised hypothesis is promptly dropped as Socrates recounts a "dream" of the perfect state ruled by wisdom (knowledge of knowledge)—the problem is that he cannot convince himself that, even in such a perfect state, everyone would really be happy. Again, the problem is the link between pure, ideal temperance or wisdom and specific happiness: what is it in knowledge of knowledge that actually makes us happy? Critias briefly suggests a compromise solution in a kind of "science of the good" (or "science of advantage"), which would apply to all other sciences but which would itself be a specific, concrete science. Socrates rejects the suggestion on the grounds that even this science of advantage does not seem to have any intelligible links to temperance as the "science of sciences."
At this point, the discussion is declared dead. Even accepting a number of things as "givens" that didn't seem very likely at the time, the two men have still been unable even to come up with a definition or a use-value for temperance. Charmides is not dissuaded by the failed discussion, however, and it is decided that he will see Socrates frequently to continue to learn from him and to pursue the true meaning of temperance.