A bet made on the flop by the player who was the aggressor before the flop, continuing the story their preflop raise began. It is the most common bet in poker, and the frequency with which people make it — and fold to it — is where a large share of small-stakes money changes hands.
It works because the preflop raiser's range contains more strong hands than the caller's, and because most flops miss most hands. If your opponent folds anything that failed to connect, a bet of half the pot needs to work only a third of the time to profit on its own, before counting the times you are actually ahead.
Board texture decides everything about it. On a dry, high-card flop where the caller's range connects poorly, a small bet can be made with almost the entire range, because there are few hands that want to fight back. On a low, connected, single-suit flop the caller's range has caught up or passed you, and betting every time turns a structural advantage into a leak. Checking there is not weakness; it is the correct answer to a board that favours the other range.
The two common failures mirror each other. Betting automatically because you raised preflop is readable within an orbit and expensive against anyone paying attention. Folding automatically to a c-bet is worse, because it makes you the reason the bet is profitable — and against an opponent who bets every flop, the correct adjustment is not to call more often with weak hands but to raise, which charges the entire bluffing range at once.