An utterance sentence generation system, which is a system that outputs an utterance sentence for responding to an utterance which is input by a user, includes a focus extraction unit that extracts focus information representing at least a portion of subject matter in the user's utterance which is input by the user, on the basis of the user's utterance, an interest state estimation unit that estimates an interest state indicating a degree of the user's interest in the subject matter represented by the focus information, a number of connected sentences determination unit that determines the number of utterance sentences to be connected, on the basis of the interest state, a connected utterance sentence generation unit that generates a connected utterance sentence by connecting utterance sentences corresponding to the number determined by the number of connected sentences determination unit, and an utterance sentence output unit that outputs the connected utterance sentence.