Listeners are able to identify spoken samples of foreign languages at better than chance levels. They are able to do this even when they do not know a language, presumably by using its general phonological properties, that is, its ``acoustic signature.'' It is not known how much exposure to a foreign language is necessary to learn to identify it. Two experiments were conducted in order to determine the ease with which American listeners could learn to identify the target language, Japanese. The first experiment examined exposure time and materials. Listeners heard word lists or a short children's story; when tested immediately after exposurse, listeners could identify Japanese interspersed with six other languages better than a control group. The materials did not affect listener performance. In the second experiment, listeners were presented the same materials but tested after a 45-min delay. This task proved much more difficult, although listener performance was marginally better than that of the control group. Apparently, listeners are able to develop a representation of the way a foreign language sounds after a brief exposure, but the representation fades rapidly.