To learn language, infants need to be able to figure out where individual words start and stop in the continuous speech streams they hear. Infants seem to be able to find words by tracking regularities in the sounds they hear in speech. This ability has been demonstrated in laboratory tasks where it is artificially quiet. To learn language however, infants need to be able to find words in speech in more real-world environments, which are typically quite noisy. The current study will investigate how infants find and learn new words from natural language input that is presented in the context of background noise.