Embodiments of the invention provide a new word identification method and apparatus. The method comprises the steps of obtaining corpora in a first preset time period before a current time point in a to-be-processed corpus library, and residual corpora, except the corpora in the first preset time period, in the to-be-processed corpus library; obtaining candidate new words of the corpora in the first preset time period according to an N-gram language model, wherein the N-gram language model represents the number of words forming the candidate new words; calculating word forming features, in the corpora in the first preset time period, of the candidate new words to obtain first word forming degrees of the candidate new words; calculating word forming features, in the residual corpora, of the candidate new words to obtain second word forming degrees of the candidate new words; according to the first word forming degrees and the second word forming degrees, calculating novel degrees of the candidate new words; and according to the first word forming degrees and the novel degrees, calculating measurement degrees of the candidate new words, sorting the measurement degrees, and identifying new words in the candidate new words. By applying the method and the apparatus, the accuracy of new word identification is effectively improved.